Controversies in Medical Physics

Controversies in Medical Physics:
a Compendium of
Point/Counterpoint Debates
Edited by:
Colin G. Orton
and
William R. Hendee
American Association of Physicists in Medicine

One Physics Ellipse
College Park, Maryland, 20740
i
Published by:
American Association of Physicists in Medicine

One Physics Ellipse
College Park, MD 20740.
Phone: (301) 209 3350 Fax: (301) 209 0862
ISBN: 978-1-888340-74-7
© Copyright by the American Association of Physicists in Medicine. All

rights reserved. No part of this book may be used or reproduced in any
manner whatsoever without written permission from the publisher, except
for brief quotations embodied in critical articles or reviews.
February, 2008
ii
TABLE OF CONTENTS
List of Contributors xi
Preface xvi
CHAPTER 1: General Radiation Therapy 1

1.1. Monte Carlo techniques should replace analytical methods for estimating dose
distributions in radiotherapy treatment planning: Radhe Mohan and John Antolak 1
1.2. Dm rather than Dw should be used in Monte Carlo treatment planning: H. Helen
Liu and Paul Keall 6
1.3. Over the next decade the success of radiation treatment planning will be judged by
the immediate biological response of tumor cells rather than by surrogate measures such
as dose maximization and uniformity: C. Clifton Ling and X. Allen Li 10
1.4. Image-guided radiotherapy is being overvalued as a clinical tool in radiation
oncology: Howard I. Amols and David A. Jaffray 15
1.5. The value of PET/CT is being over-sold as a clinical tool in radiation oncology: Lei
Xing and Barry Wessels 21
1.6. Respiratory gating for radiation therapy is not ready for prime time: X. Allen Li and
Paul J. Keall 26
1.7. Heterogeneity corrections should be used in treatment planning for lung cancer:
Nikos Papanikolaou and Eric E. Klein 32
1.8. The rate of evolution of radiation therapy planning and delivery systems is exceeding
the evolution rate of quality assurance processes: Sheri D. Henderson and Peter J. Biggs 36
1.9. Long-term changes in irradiated tissues are due principally to vascular damage in the
tissues: John Hopewell and H. Rodney Withers 40
1.10. The best radiotherapy for the treatment of prostate cancer involves
hypofractionation: John F. Fowler and Alan E. Nahum 45
1.11. The routine use of personal patient dosimeters is of little value in detecting
therapeutic misadministrations: Arnold Feldman and Fredric Marc Edwards 51
1.12. Thermoradiotherapy is underutilized for the treatment of cancer: Eduardo G. Moros
and Peter M. Corry 56
CHAPTER 2 : Highly Conformal Radiotherapy: IMRT, Tomotherapy, Stereotactic

Radiosurgery, Proton Therapy 62
2.1. It is necessary to validate each individual IMRT treatment plan before delivery:
Chester Ramsey and Scott Dube 62
2.2. Every patient receiving 3D or IMRT must have image-based target localization prior
to turning on the beam: Michael G. Herman and Douglas P. Rosenzweig 67
2.3. Linear accelerators used for IMRT should be designed as small field, high intensity,
intermediate energy units: Tiru S. Subramanian and John P. Gibbons, Jr. 72
iii
2.4. Segmental MLC is superior to dynamic MLC for IMRT delivery: Ping Xia and
Joseph Y. Ting 76
2.5. IMRT should not be administered at photon energies greater than 10 MV: David S.
Followill and Fridtjof Nüsslin 81
2.6. New radiation therapy rooms should be designed to accommodate treatments of the
future such as IMRT and tomotherapy: William G. Van de Riet and Richard G. Lane 86
2.7. IMRT may be used to excess because of its higher reimbursement from medicare:
Bhudatt R. Paliwal and Ivan A. Brezovich 90
2.8 Future developments in external beam radiotherapy will be unlikely to significantly
improve treatment outcomes over those currently achieved with 3D-conformal and IMRT
treatments: Robert J. Schulz and Dirk L. J. Verellen 95
2.9. Compared with inverse-planning, forward planning is preferred for IMRT
stereotactic radiosurgery: Fred Hacker and Daniel Low 100
2.10. Very high energy electromagnetically-scanned electron beams are an attractive
alternative to photon IMRT: Lech Papiez and Thomas Bortfeld 105
2.11. Intensity-modulated conformal radiation therapy and 3-dimensional treatment
planning will significantly reduce the need for therapeutic approaches with particles such
as protons: T. Rockwell Mackie and Alfred R. Smith 110
2.12. Proton therapy is the best radiation treatment modality for prostate cancer: Michael
F. Moyers and Jean Pouliot 114
2.13. Proton therapy is too expensive for the minimal potential improvements in outcome
claimed: Robert J. Schulz and Alfred R. Smith 119
2.14. Within the next decade conventional cyclotrons for proton radiotherapy will become
obsolete and replaced by far less expensive machines using compact laser systems for the
acceleration of the protons: Chang-Ming Charlie Ma and Richard L. Maughan 124
2.15. High energy electron beams shaped with applied magnetic fields could provide a
competitive and cost-effective alternative to proton and heavy-ion radiotherapy:
Frederick D. Becchetti and Janet M. Sisterson 129
2.16. Modern linac stereotactic radiosurgery systems have rendered the Gamma Knife
obsolete: Frank J. Bova and Steven J. Goetsch 133
2.17. In stereotactic radiosurgery, implanted fiducials are superior to an external
coordinate system: Eric G. Hendee and Wolfgang A. Tomé 137
2.18. Three-dimensional rotational angiography (3DRA) adds substantial information to
radiosurgery treatment planning of AVM'S compared to angio-CT and angio-MR: Carlo
Cavedon and Frank Bova 141
2.19. New methods for precision radiation therapy exceed biological and clinical
knowledge and institutional resources needed for implementation: Sarah S. Donaldson
and Arthur L. Boyer 146
2.20. Kilovoltage imaging is more suitable than megavoltage imaging for guiding
radiation therapy: Lei Xing and Jenghwa Chang 151
iv
CHAPTER 3: Brachytherapy 156
3.1. HDR Brachytherapy makes Cs-137 intracavitary therapy for cervix cancer a breach
of good practice: Michael T. Gillin and Jatinder R. Palta 156
3.2. Reimbursement for high dose rate brachytherapy should be based on the number of
dwell positions of the source during treatment: Geoffrey S. Ibbott and James M. Hevezi 161
3.3. Calculation of brachytherapy doses does not need TG-43 factorization: Haijun Song
and Gary Luxton 165
CHAPTER 4: General Imaging: Image Quality, Magnetic Resonance, Ultrasound 170
4.1. The development of technologies for molecular imaging should be driven principally
by biological questions to be addressed rather than by scaling down macro-level imaging
technologies: Gary D. Fullerton and John D. Hazle 170
4.2. Functional MRI is fundamentally limited by an inadequate understanding of the
origin of fMRI signals in tissue: John C. Gore and Robert W. Prost 175
4.3. Simultaneous PET/MR will replace PET/CT as the molecular multimodality imaging
platform of choice: Habib Zaidi and Osama Mawlawi 180
4.4. Methods for image segmentation should be standardized and calibrated: Edward
Chaney and Geoffrey Ibbott 185
4.5. Standards for image quality and radiation exposure impede the pursuit of optimized
quality/dose ratios in radiology: David Goodenough and Caridad Borrás 191
4.6. Diagnostic ultrasound should be performed without upper intensity limits: William
D. O'Brien, Jr. and Douglas Miller 195
CHAPTER 5: X-Ray Imaging 199
5.1. In the next decade automated computer analysis will be an accepted sole method to
separate "normal" from "abnormal" radiological images: Kenneth R. Hoffmann and Joel
E. Gray 199
5.2. In x-ray computed tomography, technique factors should be selected appropriate to
patient size: Walter Huda and Stewart C. Bushong 204
5.3. Next-generation x-ray CT units will provide <500 msec images with 3D resolution
comparable to today‘s projection radiography: Mahadevappa Mahesh and Dianna Cody 209
5.4. Free-standing, self referral centers for whole-body CT screening should be closed, or
at least restricted to at-risk patients: Richard Morin and Michael Broadbent 214
5.5. It is time to retire the computed tomography dose index (CTDI) for CT quality
assurance and dose optimization: David J. Brenner and Cynthia H. McCollough 219
5.6. A major advantage of digital imaging for general radiography is the potential for
reduced patient dose so screen/film systems should be phased out as unnecessarily
hazardous: Eliseo Vano and Keith Faulkner 224
v
5.7. Within the next five years CT colonography will make conventional colonoscopy
obsolete for colon cancer screening: Hiro Yoshida and Albert C. Svoboda, Jr. 229
5.8. For diagnostic imaging film will eventually be of historical interest only: Brent K.
Stewart and Frank N. Ranallo 235
5.9. Lossy compression should not be used in certain imaging applications such as chest
radiography: E. Russell Ritenour and Andrew D. A. Maidment 239
CHAPTER 6: Mammography 243
6.1. All mammograms should be double-read: Craig Beam and R. Edward Hendrick 243
6.2. Pre-menopausal women should be actively encouraged to seek screening
mammograms: Andrew Maidment and Elizabeth Krupinski 248
6.3. Computer-aided detection, in its present form, is not an effective aid for screening
mammography: Robert M. Nishikawa and Maria Kallergi 253
6.4. Film mammography for breast cancer screening in younger women is no longer
appropriate because of the demonstrated superiority of digital mammography for this age
group: Martin J. Yaffe and Gary T. Barnes 258
CHAPTER 7: Nuclear Medicine 264
7.1. Positron imaging with SPECT and dual-head scintillation cameras obviates the need
for PET in oncologic imaging: Mark T. Madsen and Beth A. Harkness 264
7.2. Isotopic labeling with short-lived radionuclides represents the future for clinical PET
imaging: Beth A. Harkness and Mark T. Madsen 269
7.3. Correction for image degrading factors is essential for accurate quantification of
brain function using PET: Habib Zaidi and Vesna Sossi 273
CHAPTER 8: Radiation Protection, Standards and Regulations 279
8.1. The LNT model is appropriate for the estimation of risk from low-level (less than
100 mSv/year) radiation, and low levels of radon in homes should be considered harmful
to health: Daniel J. Strom, John R. Cameron and Bernard L. Cohen 279
8.2. Exposure to residential radon causes lung cancer: R. William Field and Philippe J.
Duport 287
8.3. A prospective study should be performed to test the hypothesis that an increase in
background radiation to residents in the gulf states will increase their longevity: John R.
Cameron and Jeffrey Kahn 294
8.4. Radiation hormesis should be elevated to a position of scientific respectability: John
R. Cameron and John E. Moulder 298
vi
8.5. The use of low dose x-ray scanners for passenger screening at public transportation
terminals should require documentation of the ―informed consent‖ of passengers: Allen
F. Hrejsa and Morris L. Bank 303
8.6. The fetal dose limit for flight personnel should be 1 mSv over the gestation period,
rather than 5mSv as recommended by the NCRP: Robert J. Barish and Richard L. Morin 307
8.7. A pregnant resident physician should be excused from training rotations such as
angiography and nuclear medicine because of the potential exposure of the fetus: Edward
L. Nickoloff and Libby Brateman 312
8.8. An occupancy factor of unity should always be used for waiting rooms and other
highly-occupied public areas: Douglas R. Shearer and Michael Yester 316
8.9. States should develop regulations to require monitoring of radiation doses during
interventional fluoroscopy: Louis K. Wagner and Raymond L. Tanner 320
8.10. Reference values are de facto regulatory limits for patient exposures: Mary E.
Moore and Priscilla F. Butler 325
8.11. Federally mandated imaging standards (e.g., MQSA) serve the best interests of
patients: John McCrohan and Richard L. Morin 330
8.12. The AAPM should develop protocols generically, and avoid documents that are too
strict and prescriptive, to facilitate their adoption by regulatory agencies like the Nuclear
Regulatory Commission and States: Cynthia G. Jones and Bruce Thomadsen 334
8.13. NRC restrictions on the packaging of radioactive material should be expressed more
explicitly than simply in terms of ―apparent activity‖: Michael S. Gossman and Beth
Felinski-Semler 338
8.14. In addition to the current byproduct materials, it is important that the Nuclear
Regulatory Commission take over regulation of naturally occurring and accelerator-
produced radioactive materials: Kevin L. Nelson and J. Frank Wilson 343
8.15. The current NRC definitions of therapy misadministration are vague, do not reflect
the norms of clinical practice, and should be rewritten: Howard Amols and Jeffrey F.
Williamson 348
8.16. Medical technologies should pass performance and cost-effectiveness review in
Centers of Excellence before being released for diffusion in the clinical community:
Charles A. Kelsey and Gerald Cohen 354
CHAPTER 9: Education 358
9.1. Medical physics graduate programs should focus on education and research and
leave clinical training to residencies: Gary T. Barnes and Timothy K. Johnson 358
9.2. Physicists directing training programs should cooperate to align student enrollment
with market demand for medical physicists: Don Tolbert and James A. Zagzebski 363
9.3. All medical physicists entering the field should have a specific course on Research
and Practice Ethics in their educational background: David Switzer and Nicholas Detorie 367
vii
9.4. Physicists are better educated for a career in medical physics if they graduate from a
specialized medical physics graduate program rather than from a more traditional physics
graduate program: Ervin B. Podgorsak and David W. O. Rogers 372
9.5. Medical physics education programs should include an in-depth course in statistics
and mathematical modeling of human anatomy and physiology: Donald E. Herbert and
Gary A. Ezzell 377
9.6. Educational programs for imaging physicists should emphasize the science of
imaging rather than the technology of imaging: Paul M. DeLuca, Jr. and Mitchell M.
Goodsitt 381
9.7. Engineering is preferred over physics as an undergraduate preparation for diagnostic
and nuclear medical physicists: John D. Hazle and Charles R. Wilson 386
9.8. With the expectation that cancer detection and treatment will occur increasingly at
the molecular and gene levels, medical physics trainees should take courses in molecular
biology and genetics: Barbara Y. Croft and Colin G. Orton 390
9.9. The concepts of transient and secular equilibrium are incorrectly described in most
textbooks, and incorrectly taught to most physics students and residents: William R.
Hendee and Daniel R. Bednarek 394
9.10. Physics concepts that cannot be explained from a clinical context should be omitted
in physics courses for radiologists: G. Donald Frey and Robert L. Dixon 399
9.11. To prepare radiology residents properly for the future, their physics education
should be expanded in breadth and depth, and should be more quantitative and
mathematically-based: Michael Dennis and Mark Rzeszotarski 403
CHAPTER 10: Professional Issues 408
10.1. Medical physicists should position themselves as institutional resources in

expanding areas such as healthcare informatics and information networking: Jon H.
Trueblood and Kenneth R. Hogstrom 408
10.2. Medical physicists should assume PACS responsibilities: James R. Halama and
Walter Huda 412
10.3. It is important that medical physicists be involved in the development and
implementation of integrated hospital information systems: George C. Nikiforidis and
George C. Kagadis 416
10.4. There is no legitimate role for an applications service provider in radiology: Paul G.
Nagy and Charles E. Willis 421
10.5. The provision of consultative physics services as commodities undermines the
professionalism of medical physics: Robert Kriz and Lincoln B. Hubbard 425
10.6. The growth of biomedical engineering is a major challenge to medical physics:
Randell L. Kruger and Bruce H. Curran 429
10.7. Through their preoccupation with new technical developments, physicists have lost
sight of the realities of cancer care and statistics: Robert J. Schulz and James A. Deye 434
viii
10.8. Medical physicists would benefit by establishment of an Institute for Biomedical
Imaging: Philip F. Judy and Stephen R. Thomas 438
10.9. Candidacy for Board certification in Radiological/Medical Physics should be
restricted to graduates of accredited training programs: Timothy K. Johnson and Bhudatt
R. Paliwal 443
10.10. Graduation from a CAMPEP or equivalent accredited graduate or residency
program should be a prerequisite for certification in radiological physics: Michael G.
Herman and Howard Amols 447
10.11. Deployment of a Maintenance of Certification program in medical physics serves
the interest of the profession and the public: Stephen R. Thomas and Joseph S. Blinick 452
10.12. The proposed MOC requirements for physicists are reasonable: Stephen R.
Thomas and Jeffrey A. Garrett 457
10.13. Peer reviews of medical physics practices often yield little information because the
AAPM has not been proactive in developing appropriate peer-review guidelines: Michael
Gossman and Per Halvorsen 462
10.14. Over the foreseeable future, the growth in technical complexity of radiation
therapy will continue to drive the demand for more medical physicists: M. Saiful Huq
and Jason W. Sohn 467
10.15. The future will not need clinical therapy physicists: George Starkschall and
George W. Sherouse 472
10.16. Noncompete clauses in employment contracts violate a physicist's freedom to
practice his/her profession: Shirish K. Jani and Prakash Shrivastava 477
10.17. Medical physicists should seek patent protection for new ideas before publishing
articles about them: L. E. Antonuk and Perry Sprawls 481
10.18. A hospital-employed physicist working in radiology should provide training to
non-radiologists wishing to offer radiology services: James A. Patton and Guy Simmons 485
10.19. The Joint Commission for the Accreditation of Healthcare Organizations should
require that all physicians who perform fluoroscopy be credentialed in radiation
management by their healthcare facility: Benjamin R. Archer and David S. Gooden 489
10.20. The Ph.D. degree is a handicap in the job market for clinical medical physicists: J.
Daniel Bourland and David S. Marsden 494
10.21. Medical physicists need professional malpractice insurance: Michael Davis and
Jeffrey Masten 498
10.22. Medical physicists should not publicly endorse commercial products: Michael G.
Davis and D. Jay Freedman 502
10.23. Medical physicists should not encourage or participate in off-label, non-FDA-
approved uses of intravascular brachytherapy devices: Shirish K. Jani and Howard Ira
Amols 506
10.24. Medical Physicists should actively discourage institutions from advertising
technologies such as IMRT and HDR Brachytherapy in order to recruit patients: Marc
Edwards and William Bice 510
ix
10.25. Medical physicists in developed countries should actively help medical physicists
in developing countries: Gary D. Fullerton and Joseph Ting 515
10.26. The World Wide Web is a net disservice to medical physicists in developing
countries: Kwan-Hoong Ng and Azam Niroomand-Rad 519
10.27. Results of publicly-funded scientific research should be immediately available
without cost to the public: Colin Orton and Christopher Marshall 524
10.28. Scientific citation indices are useful in evaluating medical physicists for promotion
and tenure: David W. O. Rogers and William R. Hendee 529
10.29. Offering the physics exam to residents early in their residency is helpful to the
residents and the radiology department: Krishnadas Banerjee and Gary T. Barnes 534
10.30. Medical physicists should seek employment elsewhere when resources are
insufficient for quality patient care: Wlad T. Sobol and Ivan A. Brezovich 540
x
CONTRIBUTORS
Howard I. Amols: Memorial Sloan-Kettering Cancer Center, New York

John Antolak: Mayo Clinic, Rochester, MN
L. E. Antonuk: University of Michigan Ann Arbor, MI
Benjamin R. Archer: Baylor College of Medicine, Houston, TX
Krishnadas Banerjee: Pittsburgh, PA
Morris L. Bank: St. Vincent Hospital, Indianapolis, IN
Robert J. Barish: New York, NY
Gary T. Barnes: University of Alabama, Birmingham, AL
Craig Beam: Northwestern University Medical School, Chicago, IL
Frederick D. Becchetti: University of Michigan Ann Arbor, MI
Daniel R. Bednarek: SUNY School of Medicine, Buffalo, NY
William Bice: San Antonio, TX
Peter J. Biggs: Massachusetts General Hospital, Boston, MA
Joseph S. Blinick: Portland, ME
Caridad Borrás: Washington, DC
Thomas Bortfeld: Massachusetts General Hospital, Boston, MA
Daniel Bourland: Wake Forest University, Winston Salem, NC
Frank J. Bova: University of Florida, Gainesville, FL
Arthur L. Boyer: Scott & White Clinic, Temple, TX
Libby Brateman: University of Florida, Gainesville, FL
David J. Brenner: Columbia University, New York, NY
Ivan A. Brezovich: University of Alabama, Birmingham, AL
Michael Broadbent: Lisle, IL
Stewart C. Bushong: Baylor College of Medicine, Houston, TX
Priscilla F. Butler: American College of Radiology, Reston , VA
John R. Cameron: deceased
Carlo Cavedon: Ospedale Di Vicenza, Vicenza, Italy
Edward Chaney: University of North Carolina, Chapel Hill, NC
Jenghwa Chang: Memorial Sloan-Kettering Cancer Center, New York
Dianna Cody: U.T. M.D. Anderson Cancer Center, Houston, TX
Bernard L. Cohen: University of Pittsburgh, Pittsburgh, PA
Gerald Cohen: Milwaukee, WI
Peter M. Corry: University of Arkansas for Medical Sciences, Little Rock, AR
Barbara Y. Croft: Biomedical Imaging Program, NCI, Rockville, MD
Bruce H. Curran: University of Michigan Ann Arbor, MI
Michael Davis: The University Hospital, Cincinnati, OH
Paul M. DeLuca, Jr.: University of Wisconsin, Madison, WI
Michael Dennis: University of Toledo Medical Center, Toledo, OH
Nicholas Detorie: Sibley Memorial Hospital, Washington, DC
xi
James A. Daye: National Cancer Institute, Bethesda, MD
Robert L. Dixon: Wake Forest University, Winston Salem, NC
Sarah S. Donaldson: Stanford University Cancer Center, Stanford, CA
Scott Dube: Queen of The Valley Hospital, Napa, CA
Philippe J. Duport: University of Ottawa, Ottawa, ON, Canada
Fredric Marc Edwards: Menorah Medical Center, Overland Park, KS
Gary A. Ezzell: Mayo Clinic Scottsdale, Phoenix, AZ
Keith Faulkner: NHS/Quality Assurance Reference Centre, Wallsend, UK
Arnold Feldman: Peoria, IL
Beth Felinski-Semler: South Jersey HealthCare RMC, Vineland, NJ
R. William Field: University of Iowa, Iowa City, IA
David S. Followill: U.T. M.D. Anderson Cancer Center, Houston, TX
John F. Fowler: Medical Radiation Research Center, London, UK
D. Jay Freedman: Riverside Cancer Care Center, Newport News, VA
G. Donald Frey: Medical University of South Carolina, Charleston, SC
Gary D. Fullerton: UT Health Sciences Center, San Antonio, TX
Jeffrey A. Garrett: Mississippi Baptist Medical Center, Jackson, MS
John P. Gibbons, Jr.: Mary Bird Perkins Cancer Center, Baton Rouge, LA
Michael T. Gillin: U.T. M.D. Anderson Cancer Center, Houston, TX
Steven J. Goetsch: San Diego Medical Physics, Inc., La Jolla, CA
David S. Gooden: Saint Francis Hospital, Tulsa, Oklahoma
David Goodenough: Washington, DC
Mitchell M. Goodsitt: University of Michigan Ann Arbor, MI
John C. Gore: Vanderbilt University Medical Center, Nashville, TN
Michael S. Gossman: Tri-State Regional Cancer Center, Ashland, KY
Joel E. Gray: Medical Physics Consulting, Steger, IL
Fred Hacker: Brigham & Womens Hospital, Boston, MA
James R. Halama: Loyola University Medical Center, Maywood, IL
Per Halvorsen: Middlesex Hospital, Middletown, CT
Beth A. Harkness: Henry Ford Hospital System, Detroit, MI
John D. Hazle: U.T. M.D. Anderson Cancer Center, Houston, TX
Eric G. Hendee: Waukesha Memorial Hospital, Waukesha, WI
William R. Hendee: Medical College of Wisconsin, Milwaukee, WI
Sheri D. Henderson: Billings, MT
R. Edward Hendrick: Worthington, OH
Donald E. Herbert: University of South Alabama, Mobile, AL
Michael G. Herman: Mayo Clinic, Rochester, MN
James M. Hevezi: San Antonio, TX
Kenneth R. Hoffmann: University at Buffalo, Buffalo, NY
Kenneth R. Hogstrom: Louisiana State University, Baton Rouge, LA
John Hopewell: University of Oxford, Oxford, UK
xii
Allen F. Hrejsa: Lutheran General Hospital, Park Ridge, IL
Lincoln B. Hubbard: Hubbard, Broadbent & Assoc., Downers Grove, IL
Walter Huda: Medical University of South Carolina, Charleston, SC
M. Saiful Huq: University of Pittsburgh Medical Center, Pittsburgh, PA
Geoffrey S. Ibbott: U.T. M.D. Anderson Cancer Center, Houston, TX
Shirish K. Jani: Sharp Memorial Hospital, San Diego, CA
David A. Jaffray: Princess Margaret Hospital, Toronto, ON
Timothy K. Johnson: Rocky Mountain Cancer Center, Denver, CO
Cynthia G. Jones: USNRC, Washington, DC
Philip F. Judy: Brigham & Women's Hospital, Boston, MA
George C. Kagadis: University of Patras, Greece
Jeffrey Kahn: Minneapolis, MN
Maria Kallergi: H. Lee Moffitt Cancer Center, Tampa, FL
Paul Keall: Stanford University Cancer Center, Stanford, CA
Charles A. Kelsey: University of New Mexico, Albuquerque, NM
Eric E. Klein: Washington University, Saint Louis, MO
Robert Kriz: Tinley Park, IL
Randell L. Kruger: Marshfield Clinic, Marshfield, WI
Elizabeth Krupinski: University of Arizona, Tucson, AZ
Richard G. Lane: Houston, TX
X. Allen Li: Medical College of Wisconsin, Milwaukee, WI
C. Clifton Ling: Memorial Sloan-Kettering Cancer Center, New York
H. Helen Liu: U.T. M.D. Anderson Cancer Center, Houston, TX
Daniel Low: Washington University, Saint Louis, MO
Gary Luxton: Stanford University School of Medicine, Stanford, CA
Chang-Ming Charlie Ma: Fox Chase Cancer Center, Philadelphia, PA
T. Rockwell Mackie: University of Wisconsin – ADCL, Madison, WI
Mark T. Madsen: University of Iowa, Iowa City, IA
Mahadevappa Mahesh: Johns Hopkins University, Baltimore, MD
Andrew D. A. Maidment: University of Pennsylvania, Philadelphia, PA
David S. Marsden: Marsden Medical Physics Associates, Kinnelon, NJ
Christopher Marshall: NYU Medical Center, New York, NY
Jeffrey Masten: Rapid City Regional Hospital, Rapid City, SD
Richard L. Maughan: University of Pennsylvania, Philadelphia, PA
Osama Mawlawi: U.T. M.D. Anderson Cancer Center, Houston, TX
Cynthia H. McCollough: Mayo Clinic, Rochester, MN
John McCrohan: CDRH, Rockville, MD
Douglas Miller: University of Michigan, Ann Arbor, MI
Radhe Mohan: U.T. M.D. Anderson Cancer Center, Houston, TX
Mary E. Moore: Philadelphia VA Medical Ctr., Philadelphia, PA
Richard Morin: Mayo Clinic, Jacksonville, FL
xiii
Eduardo G. Moros: University of Arkansas for Medical Sci., Little Rock, AR
John E. Moulder: Medical College of Wisconsin, Milwaukee, WI
Michael F. Moyers: Proton Therapy, Inc., Marlton, New Jersey
Paul G. Nagy: University of Maryland Medical Center, Baltimore, MD
Alan E. Nahum: Clatterbridge Centre for Oncology, Wirral, Merseyside, UK
Kevin L. Nelson: Mayo Clinic, Jacksonville, FL
Kwan-Hoong Ng: University of Malaya, Kuala Lumpur, Malaysia
Edward L. Nickoloff: Columbia University, New York, NY
George C. Nikiforidis: University of Patras, Greece
Azam Niroomand-Rad: Madison, WI
Robert M. Nishikawa: University of Chicago, Chicago, IL
Fridtjof Nüsslin: Klinik für Strahlentherapie, Munich, Germany
William D. O'Brien, Jr.: University of Illinois, Urbana, IL
Colin G. Orton: Wayne State University, Detroit, MI
Bhudatt R. Paliwal: University of Wisconsin, Madison, WI
Jatinder R. Palta: University of Florida Health Science Center, Gainesville, FL
Nikos Papanikolaou: Cancer Therapy & Research Center, San Antonio, TX
Lech Papiez: UT Southwestern Medical Center, Dallas, TX
James A. Patton: Vanderbilt University Medical Center, Nashville, TN
Ervin B. Podgorsak: McGill University Health Center, Montreal, QC, Canada
Jean Pouliot: UC San Francisco, San Francisco, CA
Robert W. Prost: Froedtert Hospital, Milwaukee, WI
Chester Ramsey: Thompson Cancer Survival Center, Knoxville, TN
Frank N. Ranallo: University of Wisconsin, Madison, WI
E. Russell Ritenour: University of Minnesota Medical School, Minneapolis, MN
David W. O. Rogers: Carleton University, Ottawa, ON, Canada
Douglas P. Rosenzweig: University of Rochester Medical Center, Rochester, NY
Mark Rzeszotarski: MetroHealth Medical Center, Cleveland, OH
Robert J. Schulz: Johnson, VT
Douglas R. Shearer: Providence, RI
George W. Sherouse: Sherouse Systems, Inc., Chapel Hill, NC
Prakash Shrivastava: University of Southern California, Los Angeles, CA
Guy Simmons: Bluegrass Radiological Physics, Lexington, KY
Janet M. Sisterson: Massachusetts General Hospital, Boston, MA
Alfred R. Smith: U.T. M.D. Anderson Cancer Center, Houston, TX
Wlad T. Sobol: University of Alabama Birmingham, Birmingham, AL
Jason W. Sohn: Case Western University, Cleveland, OH
Haijun Song: Duke University Medical Center, Durham, NC
Vesna Sossi: University of Columbia, Vancouver, BC, Canada
Perry Sprawls: Sprawls Educational Foundation, Montreat, NC
George Starkschall: U.T. M.D. Anderson Cancer Center, Houston, TX
xiv
Brent K. Stewart: University of Washington, Seattle, WA
Daniel J. Strom: Pacific Northwest National Laboratory, Richland, WA
Tiru S. Subramanian: North Bay Cancer Center, Fairfield, CA
Albert C. Svoboda, Jr.: Santa Barbara, CA
David Switzer: Northern Rockies Cancer Center, Billings, MT
Raymond L. Tanner: Memphis, TN
Bruce Thomadsen: University of Wisconsin, Madison, WI
Stephen R. Thomas: University of Cincinnati Medical Center, Cincinnati, OH
Joseph Y. Ting: Melbourne Cancer Center, Melbourne, FL
Don Tolbert: Beloit, KS
Wolfgang A. Tomé: University of Wisconsin, Madison, WI
Jon H. Trueblood: Medical College of Georgia, Augusta, GA
William G. Van de Riet: Bloomington Hospital, Bloomington, IN
Eliseo Vano: Complutense University, Madrid, Spain
Dirk L. J. Verellen: AZ-VUB Radiotherapy, Brussels, Belgium
Louis K. Wagner: UT Medical School, Houston, TX
Barry Wessels: Univ Hospitals-Case Western Reserve University, Cleveland, OH
Jeffrey F. Williamson: Virginia Commonwealth University, Richmond, VA
Charles E. Willis: U.T. M.D. Anderson Cancer Center, Houston, TX
Charles R. Wilson: Medical College of Wisconsin, Milwaukee, WI
J. Frank Wilson: Medical College of Wisconsin, Milwaukee, WI
H. Rodney Withers: UCLA Medical Center, Los Angeles, CA
Ping Xia: UC San Francisco, San Francisco, CA
Lei Xing: Stanford University School of Medicine, Stanford, CA
Martin J. Yaffe: Sunnybrook HSC, Imaging Research, Toronto, ON, Canada
Michael Yester: UAB Medical Center, Birmingham, AL
Hiro Yoshida: Harvard Medical School, Boston, MA
James A. Zagzebski: University of Wisconsin, Madison, WI
Habib Zaidi: Geneva University Hospital, Geneva, Switzerland
xv
PREFACE
The Point/Counterpoint series of debates in Medical Physics began in March 1998

and has continued unabated since. Point/Counterpoints are among the most popular
articles read in Medical Physics as demonstrated by consistently high online readership
statistics. Indeed, they are often the most downloaded of all articles in the monthly
statistics. To commemorate the first 10 years of Point/Counterpoint debates and,
coincidentally, the first 50 years of existence of the American Association of Physicists in
Medicine, the journal‘s Editorial Board decided that to publish a compendium of the
debates as a separate, free-access, online book with the title ―Controversies in Medical
Physics‖.
Although the Point/Counterpoints included in this first volume of debates have
been reformatted, they are essentially identical to those that appeared in the journal with
one exception—the online version contains links to references within the text and to
references cited by the authors. Readers will need to access the original articles in the
online journal [http://scitation.aip.org/journals/doc/MPHYA6-home] to take advantage of
these citation links. Each Point/Counterpoint has a link on the title page to the original
online Abstract where readers can access the full articles if they or their institutions are
subscribers to the journal.
Point/Counterpoint debates included in this first volume were moderated by either
Bill Hendee (1998–2005) or Colin Orton (2005–2007). The moderators devised most of
the propositions, selected appropriate authors and edited their contributions. Persons
participating in Point/Counterpoint discussions were selected for their knowledge and
communicative skills, and a disclaimer preceded all Point/Counterpoints to the effect that
the positions of the authors for or against a proposition ―may or may not reflect their
personal opinions or the positions of their employers.‖
We hope you enjoy reading the Point/Counterpoint debates included in this first
online volume, and look forward to suggestions you may have for future volumes in the
series.
Colin G. Orton & William R. Hendee

Editors
February, 2008
xvi
1
CHAPTER 1
General Radiation Therapy
1.1. Monte Carlo techniques should replace analytical methods for

estimating dose distributions in radiotherapy treatment planning
Radhe Mohan and John Antolak

Reproduced from Medical Physics, Vol. 28, No. 2, pp. 123–126, February 2001
(http://scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=MPHYA6000028000002
000123000001&idtype=cvips&gifs=Yes)
OVERVIEW
Analytical models have traditionally been used to estimate dose distributions for treatment
planning in radiation therapy. Recently, some physicists have suggested that Monte Carlo
techniques yield more accurate computations of dose distributions, and a few vendors of
treatment planning systems have incorporated Monte Carlo methods into their software. Other
physicists argue that, for a number of reasons, analytical methods should be preserved. This
controversy is the topic of this Point/Counterpoint article. Thanks are extended to Paul Nizin,
Ph.D. of Baylor College of Medicine for suggesting the topic.
Arguing for the Proposition is Radhe Mohan, Ph.D. Dr. Mohan received his Ph.D. from Duke
University and is currently Professor and Director of Radiation Physics at the Medical College of
Virginia (MCV) Hospitals, Virginia Commonwealth University. Dr. Mohan has been actively
engaged in research and clinical implementation of advanced dose calculation methods, 3D
treatment planning, Monte Carlo techniques and IMRT for 25 years, first at Memorial Sloan-
Kettering Cancer Center and now at MCV. He has published and lectured widely on these topics
at national and international meetings and symposia.
Arguing against the proposition is John Antolak, Ph.D. Dr. Antolak received his Ph.D. in
Medical Physics from the University of Alberta (Canada) in 1992. He then joined the Department
of Radiation Physics at The University of Texas M. D. Anderson Cancer, where he is currently
an Assistant Professor. He is certified by the American Board of Radiology and licensed to
practice Medical Physics in Texas. He is active in the education of graduate students,
dosimetrists, and other physicists, and his research interests center around the use of electron
beams for conformal radiotherapy. In his spare time, he enjoys playing ice hockey and coaching
his son's ice hockey team.
FOR THE PROPOSITION: Radhe Mohan, Ph.D.
Opening Statement
2
Monte Carlo techniques produce more accurate estimates of dose than other computational
methods currently used for planning radiation treatments. Were it not for limitations of computer
speed, Monte Carlo methods probably would have been used all along.
With the spectacular increase in computer speed and the development of clever algorithms and
variance reduction schemes, Monte Carlo methods are now practical for clinical use. The time
required to compute a typical treatment plan has shrunk to a few minutes on computers costing
less than $50,000. A few centers have started using Monte Carlo techniques for clinical purposes,
and releases of commercial products are imminent.
As with any new product, an "adjustment period" will be needed during which we learn how to
apply this powerful tool. Some find the "statistical jitter" in Monte Carlo results troubling. This
issue is being addressed by several investigators. The additional cost of hardware and software
may be another obstacle, but is likely to be resolved as computers become cheaper and more
powerful.
Another issue is whether improvements in accuracy are clinically significant and worth the
additional cost. It is difficult to answer the first question unequivocally, because randomized
trials in which half the patients are treated with less accurate methods are not feasible. Arguments
in favor of using Monte Carlo methods include:
(1) Elimination of the need to continually reinvent approximate dose computation models and to
tweak them to meet every new situation, as well as the need for trial and error approaches to
obtain acceptable matches with measured data. The medical physics community has been
engaged in such exercises for 50 years. It is time to stop.
(2) Broad applicability and accuracy of the same Monte Carlo model for all anatomic geometries
and treatment modalities (photons, electrons, brachytherapy). With analytical methods, there is a
separate model for each modality and a unique set of approximations and assumptions is required
for each type of field shaping device.
(3) Dramatic reduction in the time, effort and data required for commissioning and validating the
dose computation part of treatment planning systems.
(4) Improved consistency of inter-institutional results, and greater quality of dose response data
because of improved dose accuracy.
(5) Accurate estimation of quantities difficult or impossible to measure, such as dose distributions
in regions of disequilibrium.
Until recently, the major reason for considering Monte Carlo methods was the inaccuracy of
semi-empirical models for internal inhomogeneities and surface irregularities. Now an equally
important justification is the ability of Monte Carlo techniques to provide accurate corrections for
transmission through, scattering from, and beam hardening by field shaping devices. Monte
Carlo techniques are also able to account correctly for radiation scattered upstream from field-
shaping devices. These effects are quite significant for small fields encountered in intensity-
modulated radiotherapy.
The transition to Monte Carlo methods will have to be gradual. Even though a few minutes of
time to compute a plan may seem insignificant, computer-aided optimization of treatment plans
may require many iterations of dose computations. In these situations, hybrid techniques will be
needed that use fast but less accurate conventional models for most optimization iterations and
Monte Carlo techniques for the remainder.
Since Monte Carlo techniques are now affordable and practical, there is no reason not to use
them. It is not necessary to conduct clinical trials to once again prove the clinical significance of
3
improved dose accuracy. Monte Carlo methods should be deployed in radiation therapy with
deliberate speed. For some applications, such as, IMRT optimization, it may be necessary to
continue to use high-speed conventional methods in conjunction with Monte Carlo techniques at
least for now.
Rebuttal
Dr. Antolak has raised several issues, some of which were addressed in my Opening Statement.
With faster computers and clever schemes to reduce variance, the stochastic nature of the Monte
Carlo approach is no longer an impediment. Statistical uncertainty of 1%–2% is achievable on
grid sizes of 2–3 mm in MC dose distribution calculations, requiring just a few minutes on easily
affordable multiprocessor systems. While statistical noise may be unsightly, its effect on the
evaluation of dose-volume and dose-response parameters of plans is insignificant. In addition,
techniques to smooth out noise are being implemented.
Analytic models introduce systematic errors in dose. They simply cannot achieve the accuracy of
MC techniques. While it is true that analytic models consistently produce precise results for the
same input data, these results are consistently inaccurate.
Dr. Antolak is concerned that approximations to speed up Monte Carlo computations may affect
the accuracy of results. But Monte Carlo developers and users should always ensure that
approximations have no significant impact on accuracy. Nothing else should be necessary.
Responses to other such concerns raised by Dr. Antolak are: (1) Considering the uncertainties in
dose-response information and other sources of data in the radiotherapy chain, our ability to
define "how much noise in the dose distributions is acceptable" is similar to our ability (or lack
thereof) to determine the level of dose inaccuracy that may be acceptable. (2) Dose to a point is
not a meaningful quantity when Monte Carlo techniques are used. Beam weighting and dose
prescription should be specified in terms of dose to fractional volumes (e.g., 98% of the tumor
volume). (3) Statistical noise should have practically no effect on inverse treatment planning
because the intensity along a ray is affected by the average of dose values over a large number of
voxels lying along the ray and not by the dose in any one voxel. (4) Commissioning of Monte
Carlo algorithms will be the responsibility of the same physicists and/or commercial vendors who
commission conventional methods.
I believe strongly that concerns raised by Dr. Antolak and others are being resolved and that we
are now ready to introduce Monte Carlo techniques into clinical use.
AGAINST THE PROPOSITION: John Antolak, Ph.D.
Opening Statement
We have a professional responsibility to ensure that patient treatments are accurately delivered,
and the accuracy of treatment planning dose computation is one aspect of this. There are data to
support the conclusion "that Monte Carlo techniques yield more accurate computations of dose
distributions," provided that the Monte Carlo technique is fully applied. However, in light of
other factors detailed below, Monte Carlo methods should not replace analytical methods for
estimating dose distributions.
4
Before arguing against the proposition, it is necessary for me to clarify what I believe the basic
difference is between an analytical method and a Monte Carlo method. It boils down to the
difference between deterministic and stochastic. The Monte Carlo method is stochastic, i.e.,
independent calculations of the same problem will give different answers. The analytical method
is deterministic, i.e., independent calculations of the same problem will give the same answer, at
least to within the limits of numerical round-off and truncation errors. In my opinion and for the
purpose of this discussion, any nonstochastic method is considered to be an analytical method.
The accuracy of an algorithm (or method) describes how close it comes to the true answer.
Clinical physicists have to worry about the accuracy of both analytical and Monte Carlo methods.
The full Monte Carlo method (e.g., EGS4) is considered by many to be the gold standard for
accurate dose calculations. The precision of an algorithm is a measure of the repeatability of the
answer. Analytical methods have essentially absolute precision. However, the precision of the
Monte Carlo method, as measured by the standard error, is proportional to the inverse of the
volume of the dose voxels, and to the inverse square root of the computational resources
allocated to the problem. For example, reducing the standard error by a factor of two requires
four times as much CPU-time. Variance reduction techniques can be used to reduce the
computational resources required to obtain a given precision. However, the time (or resources)
required for full Monte Carlo simulations of patient dose distributions is currently too great for
clinical use. By necessity, current Monte Carlo treatment planning algorithms (those being touted
for clinical use) introduce approximations that greatly speed up the calculations, but the accuracy
of the results may be affected. At the same time, significant improvements are also being made to
the accuracy of analytical algorithms. Also, for the clinical physicist, commissioning analytical
algorithms is relatively straightforward, noise is not a problem, and the accuracy can be easily
documented.
From the perspective of the clinical physicist, many questions about the use of Monte Carlo
algorithms have not yet been answered. How much noise in the dose distributions is acceptable?
In the presence of noise, how should beam weighting (e.g., isocentric weighting) be done? What
effect does noise have on inverse treatment planning? Who will take responsibility for
commissioning the algorithm, and how accurate are the results of the commissioning? How long
will the calculation take relative to faster analytical calculations? How will the calculation time
affect treatment-planning throughput, particularly when using optimization methods? Is the
spatial resolution sufficient for clinical use? Most of the time, Monte Carlo treatment planning
calculation times are quoted for relatively coarse (e.g., 5 mm) spatial resolution. Just reducing the
resolution from 5 mm to 3 mm requires approximately five times as much CPU-time. These are
just some of the issues that need to be resolved and well-documented before Monte Carlo
methods can replace analytical methods for treatment planning.
Monte Carlo methods may be used as an independent verification of the dose delivery, or to
document (rather than plan) the dose delivery. However, until the questions above are
successfully answered, Monte Carlo methods should not replace analytical methods for
estimating radiotherapy dose distributions.
Rebuttal
5
We agree that greater accuracy in dose computation is desirable, Monte Carlo methods can
produce more accurate dose estimates, and "Monte Carlo methods should be deployed in
radiation therapy with deliberate speed." However, these points are not the proposition we are
addressing.
A potential patient recently inquired about the status of Monte Carlo planning at our institution.
From what he had read, he believed that Monte Carlo treatment planning is a "silver bullet." Dr.
Mohan says it is time to stop reinventing. I believe that implementation of "clever algorithms and
variance reduction schemes" is reinventing Monte Carlo treatment planning methods. Further,
trial and error will not stop with Monte Carlo. With complete information about source and
machine geometry, Monte Carlo calculations can be highly accurate. However, Monte Carlo
algorithms usually start with a source model that requires trial and error adjustments to match
measured data.
Whereas reinventing analytical methods usually improves accuracy, reinventing Monte Carlo
methods may decrease accuracy. In both cases, there is a tradeoff between accuracy and speed,
which is often seen if the Monte Carlo approach averages the dose over large voxels. How is the
accuracy of a particular implementation judged and to what should it be compared? The
"spectacular increase in computer speed and the development of clever algorithms" noted by Dr.
Mohan permits significant improvements in analytical models, potentially leading to a model for
coupled photon-electron transport under more general conditions.
Future reductions in commissioning and validation efforts will only come from manufacturers'
standardization of treatment machines and improved quality in their construction. Modified and
new machines will still require extensive commissioning and validation for both Monte Carlo and
analytical methods. Considerable research remains to be done to identify the minimum data set
sufficient to validate input data that characterizes a treatment machine. Dr. Mohan's last two
points are really arguments for greater dose accuracy and apply to treatment planning systems in
general, not just Monte Carlo methods.
Dr. Mohan cites the significance of Monte Carlo methods applied to field shaping and intensity
modulation devices. These applications can be very complex, but are usually not modeled
explicitly. For example, modeling all of the field segments for a complex DMLC fluence pattern
is impractical under normal circumstances. Using an approximate approach affects the overall
accuracy of the dose calculation, as it would for an analytical method.
Monte Carlo methods are an invaluable tool for improving analytical models to a point where
their dose uncertainty is insignificant compared with other uncertainties radiation therapy—such
as setup, internal organ motion, target delineation, and biological response. As stated earlier,
"Monte Carlo methods may be used as an independent verification of dose delivery, or to
document (rather than plan) dose delivery," but should not replace analytical methods for
estimating dose distributions in radiotherapy treatment planning.
6
1.2. Dm rather than Dw should be used in Monte Carlo treatment

planning
H. Helen Liu and Paul Keall

Reproduced from Medical Physics, Vol. 29, No. 5, pp. 922–924, May 2002
OVERVIEW
Dw, the absorbed dose to water, has traditionally been the normalizing factor for dose
computations related to treatment planning in radiation therapy with high-energy x-rays. This
normalizing factor relates the treatment plan to the x-ray calibration process described in TG 21
or 51. However, treatment planning employing Monte Carlo techniques allows the expression of
radiation transport and energy deposition in patient representative media. The dose reported in
this process can be either the dose-to-medium, Dm, or the dose-to-water, Dw, calculated with
stopping power ratios. Many physicists believe that Dm is the preferred variable for treatment
planning, and that it should replace Dw in this capacity. The controversy of Dm versus Dw is the
topic of this month's Point/Counterpoint.
Arguing for the Proposition is H. Helen Liu, Ph.D. Dr. Liu worked on radiation dosimetry and
Monte Carlo simulation for her Ph.D. research in the Division of Radiation Oncology at the
Mayo Clinic. She completed residency training at the Mayo Clinic upon completion of her Ph.D.
in 1997. She is now an Assistant Professor in the Department of Radiation Physics at the
University of Texas MD Anderson Cancer Center. Her research interests include Monte Carlo
simulation, treatment planning optimization, and biophysical models for measuring radiation
response in radiation therapy.
Arguing against the Proposition is Paul Keall, Ph.D. Dr. Keall is currently an Assistant Professor
in the Department of Radiation Oncology at the Medical College of Virginia. He has been
working on Monte Carlo treatment planning-related research for over a decade. His Ph.D.
dissertation involved the development and evaluation of "Superposition Monte Carlo," a dose
calculation algorithm combining elements of both the superposition/convolution and Monte
Carlo algorithms. Paul has acted as a consultant to the IAEA on Monte Carlo Transport in
Radiotherapy. His current Monte Carlo research interests include hip prosthesis calculations,
IMRT calculations, EPID dosimetry and large-scale comparisons between Monte Carlo and other
algorithms.
FOR THE PROPOSITION: H. Helen Liu, Ph.D.
Opening Statement
The ability to compute the actual dose to medium (Dm) is a unique and advantageous feature of
Monte Carlo simulation for radiotherapy treatment planning. The rationale of converting Dm back
to the dose to water Dw is driven solely by the desire to comply with tradition. Dw has been used
in treatment planning because accurate heterogeneity correction methods were not readily
available. As Monte Carlo treatment planning emerges, new standards of practice will be
established to reflect the advances that Monte Carlo techniques will bring. The motivation for
7
using Monte Carlo simulation in treatment planning is essentially to achieve the greatest
accuracy in dose calculation. Converting Dm back to Dw requires computing stopping power
ratios for local voxels, a process that adds uncertainty in the calculations and makes Monte Carlo
simulation more time consuming and complicated. This conversion defeats the purpose of using
Monte Carlo simulation.
The clinical impact of switching from Dw to Dm is not expected to be significant, mainly because
most tissues of interest in radiotherapy are similar to water. Thus, the difference between Dm and
Dw will not change clinical outcomes to a noticeable degree, particularly since the uncertainty of
clinical results is several orders of magnitude greater. With respect to radiobiological effects,
there is no reason that Dm cannot be used in place of Dw for purposes of linking dose to biological
response. In fact, the energy deposited in individual voxels is related more directly to Dm than to
Dw.
Insofar as the dosimetry calibration protocol is concerned, the use of Dm in treatment planning
will not affect the calibration of Dw recommended by national and international protocols. This is
because in Monte Carlo simulation, the relationship is known between dose (either Dw or Dm) and
the required number of photon histories to be simulated. From a prescription of Dm in a patient,
the corresponding number of photon histories can be converted to monitor units (MUs) through
use of the calibrated value of Dw per MU. In other words, the calculation of MU can proceed in
the same way for either Dw or Dm in Monte Carlo simulation.
In summary, the advantages of Monte Carlo planning systems are improved accuracy in dose
calculations and the possibility of obtaining Dm directly for various tissues. Converting Dm to Dw
involves additional complications and adds possible sources of error for Monte Carlo
calculations. Dm should be used in Monte Carlo planning, and will not have a significant impact
on current clinical practice. Instead, use of Dm allows Monte Carlo planning to establish more
accurate dose delivery, and to provide a closer relationship between tissue response and dose.
Rebuttal
Dr. Keall raises some important issues concerning the use of Monte Carlo treatment planning in
routine clinical practice. Included in his concerns is a preference for Dw rather than Dm in Monte
Carlo treatment planning.
One solution to Dr. Keall point (3) on specification of the medium is to standardize the
conversion of CT numbers to radiological properties of the medium in a consistent and uniform
manner for different Monte Carlo treatment planning systems. I suspect that this effect may not
be clinically significant because Dw and Dm are quite similar for most biological tissues.
Nevertheless, the subject warrants further investigation.
With respect to the dosimetry calibration protocol, the relationship between Dw obtained in a
reference calibration condition, and Dm for a patient prescription, is known from Monte Carlo
simulation. Hence, the monitor unit calculation in Monte Carlo treatment planning is
straightforward, and does not affect implementation of standard dosimetry protocols to any
degree. For conventional treatment planning, many institutions use dose to muscle rather than
dose to water for monitor unit calculation. The ratio between the two is well-known and has been
used to scale monitor units in clinical practice.
Obstacles to using Dm rather than Dw for Monte Carlo treatment planning are pragmatic in nature,
including how to conform with convention and incorporate past clinical experience. The
8
relationship between Dw and Dm can be computed, and dose response data for tumor and normal
tissues can be easily scaled in order to use Dm for treatment planning purposes. In fact, for cells
imbedded in heterogeneous tissues such as lung or bone, dose to the cells is not reflected
accurately by either Dw or Dm. This is because the CT imaging resolution is not sufficient to
detect subvoxel structures, and Dw or Dm simply represents an averaged dose value in the voxel.
In this case, either Dw or Dm can be used to indicate the energy delivered to the cells and the
subsequent radiobiological effects.
The use of Dm in Monte Carlo treatment planning is a natural and suitable approach to avoiding
the additional complexity and uncertainty of converting Dm to Dw. New standards of practice
using Dm should be implemented to provide a smooth transition from conventional to Monte
Carlo treatment planning.
AGAINST THE PROPOSITION: Paul Keall, Ph.D.
Opening Statement
I aim to convince you that prescribing, evaluating and reporting dose-to-water (Dw) rather than
dose-to-medium (Dm) for Monte Carlo treatment planning is both obvious and necessary for the
following reasons: (1) Clinical experience is Dw-based, (2) Dosimetry protocols are Dw-based, (3)
The "medium" to report dose in is always a guess, and (4) Dw-based IMRT allows us to achieve
the clinical prescription.
1. Clinical experience is Dw-based. Since the introduction of computer-based treatment

planning in the 1960s, dose calculation algorithms have assumed that the patient is composed of
waterlike composition of varying density. All manual calculations assume waterlike composition.
This assumption is reasonable, as water makes up the bulk of our body cells and tissues. All
clinical experience, and doses reported in the multitude of clinical trials, both past and ongoing,
are with respect to Dw.
2. Dosimetry protocols are Dw-based. As stated in the overview, modern dosimetry

protocols (e.g., IAEA, AAPM) are based on Dw. Also, the factors that convert ionization-in-air to
Dw for the calibration have been determined or verified by Monte Carlo calculations. It seems
reasonable that the reported dose for treatment planning is directly traceable to the calibration,
and thus the reported dose should be Dw.
3. The "medium" to report dose in is always a guess. In Monte Carlo treatment planning, the
CT numbers are converted not only to densities, but also to media. These media are generally
obtained from ICRU or ICRP publications. However, who knows whether a patient's organ, e.g.,
a liver, has exactly the same composition as the `standard" liver? Furthermore, the same CT
number can be obtained from a higher density/lower Z medium (e.g., soft tissue) and from a
lower density/higher Z medium (e.g., lung). Thus, in the absence of 3D body composition
analysis accompanying the CT scan, we can only guess at patient composition. Errors in
specifying the medium for the dose calculation will always be present. Thus, reporting Dm when
we do not know what the medium is seems somewhat illogical.
4. Dw-based IMRT allows us to achieve the clinical prescription. The prescription dose is
determined with the aim that the tumor cells receive a lethal dose, whilst the normal cells
embedded in the tumor receive a dose they can recover from. Take, for example, a head and neck
9
cancer in which the tumor infiltrates the mandible. If optimized using Dm, the dose-to-mandible
will be equal to the prescription dose, say 72 Gy. However, the dose-to-osteocytes (being
waterlike) within the mandibular bony matrix will be close to 80 Gy (assuming a Bragg-Gray
cavity). Hence there will be a higher chance of bone weakening, fracture and necrosis than at 72
Gy. If optimized using Dw, we can modulate the intensity and thus dose so that the dose-to-tumor
(being waterlike) and dose-to-osteocytes within the mandible receive the prescription dose, 72
Gy. The point here is that Dw, rather than Dm, is the desired quantity from a clinical perspective,
and IMRT with Dw Monte Carlo can closely approach the desired prescription.
From the four reasons mentioned above, I conclude that we should use Dw rather than Dm for
Monte Carlo treatment planning (as we routinely use in our clinic for our IMRT patients).
Furthermore I would advise both commercial and academic Monte Carlo developers to report Dw
rather than Dm.
Rebuttal
Dr. Liu presents arguments that converting Dm to Dw introduces (i) additional complexity and (ii)
additional uncertainty.
On the issue of increased complexity. Converting Dm to Dw does involve additional computations.

There is a small computational overhead associated with scoring Dw rather than Dm. However, we
must ask ourselves why we want Monte Carlo in the first place? If complexity was a limitation,
we would still be performing 2D treatment planning. If we make the effort to implement Monte
Carlo treatment planning, let us do it properly.
On the issue of uncertainty. The additional uncertainty that Dr. Liu is alluding to is the
uncertainty of the stopping power ratios of water-to-medium. By virtue of taking a ratio,
systematic uncertainties in the stopping powers are reduced. Furthermore, the stopping power
values for water are more accurate than those for many other tissues. There are two uncertainties,
however, that exceed those of stopping power values. These are (i) the CT voxel composition
uncertainty and (ii) institutional variations.
(i) A 1×1×3 mm3 CT voxel can contain up to 300 thousand cells of differing types and
composition. Even the most accurate Monte Carlo is limited by the CT resolution. When we talk
about scoring Dm to a voxel, which cell type do we mean? Assuming cells are Bragg-Gray
cavities, each cell type within the CT voxel will have a different Dm. However, when we score
Dw, Bragg–Gray theory tells us that the dose to each cell within the CT voxel is the same, thus
eliminating intra-CT voxel variations.
(ii) Different institutions and Monte Carlo developers/vendors use different material types
and density cut offs. For example, the default number of patient representative media in
DOSXYZ (a widely used EGS4 Monte Carlo user code) is four. However, advanced users are
including either quasicontinuous or continuous media assignments. Thus, Dm will be site
dependent. At least, by virtue of Bragg–Gray theory, if Dw is used institutional variations will be
reduced.
Dr. Liu also states that the rationale of converting Dm to Dw is driven simply by the need to
comply with convention. I completely agree. As with dose calibration, for clinical trial results to
be meaningful, the dose reported by different institutions should have a consistent convention.
10
1.3. Over the next decade the success of radiation treatment

planning will be judged by the immediate biological response of
tumor cells rather than by surrogate measures such as dose
maximization and uniformity
C. Clifton Ling and X. Allen Li

Reproduced from Medical Physics, Vol. 32, No. 7, pp. 2189–2192, July 2005
002189000001&idtype=cvips&gifs=Yes )
OVERVIEW
Radiation treatment plans have traditionally been optimized by surrogate measures such as dose
maximization and uniformity in the target and dose limits to normal tissues. But with emerging
molecular technologies, the possibility is evolving to monitor biological responsiveness directly
in real-time during treatment with radiation. Whether this approach will replace traditional dose-
optimization approaches is the subject of this month's Point/Counterpoint.
Arguing for the Proposition is C. Clifton Ling, Ph.D. Dr. Ling received his Ph.D. in nuclear
physics from the University of Washington, Seattle. He has performed studies in radiation and
medical physics at Memorial Sloan-Kettering (MSKCC), Massachusetts General, George
Washington University, and the University of California San Francisco. He is the Enid Haupt
Professor and Chair of Medical Physics, MSKCC. Dr. Ling's research interests range from
molecular radiation biology to IMRT, and recently biological imaging for cancer diagnosis,
characterization and treatment. Dr. Ling has authored 200 peer-reviewed papers and numerous
book chapters.
Arguing against the Proposition is X. Allen Li, Ph.D. Dr. Li is an Associate Professor and the
Chief Physicist in the Department of Radiation Oncology, Medical College of Wisconsin. He has
been a board certified medical physicist since 1994, and is currently a member of AAPM TG-74
and TG-106. Dr. Li is an Associate Editor for Medical Physics and has published 82 peer-
reviewed papers. His research interests include biologically-based treatment planning and image-
guided radiation therapy.
FOR THE PROPOSITION: C. Clifton Ling, Ph.D.
Opening Statement
Deficiencies in current radiotherapy planning, directly linked to the history of radiotherapy, are
due to the paucity and/or inadequate understanding of biological and clinical data needed for
optimum treatment design. Treatment planning was nonexistent or primitive prior to the
application of computers in the 1960–70s, and after that, primarily by "convention" such as
parallel opposed fields, four-field box or arc-therapy. As to biological and clinical issues, little
was known about tumor or organ dose response, dose-volume relationships, and the
11
heterogeneities of tumor burdens and of radiosensitivity within the tumors. In the absence of
meaningful criteria, dose-based surrogates were used by default.
Since the 1980s, 3-dimensional conformal radiotherapy (3D-CRT) technology has made
available complete dose-volume information which, when combined with clinical outcome data,
has permitted model-based analyses of tumor local control and organ toxicity. These analyses
clearly reveal that dose-surrogates by themselves are insufficient to predict the consequences of a
particular treatment plan. For example, the tumor control probability of prostate cancers depends
on the risk groups in which the patients belong,1 lung toxicities are more likely in the lower than
in the upper lobe,2 and rectal bleeding may depend on functional reserve. 3 All biophysical
models that attempt to describe clinical outcomes contain biological information, be it SF2
(cellular surviving fraction for a dose of 2 Gy) in the Equivalent-Uniform Dose (EUD) model,4
or the architecture of an organ's functional subunits.5
The success of the human genome project and developments in imaging strongly suggest that
biological data will be increasingly available for radiotherapy planning in the future. The
influence of various gene products on the radiosensitivity and chemosensitivity of tumors and
normal tissues is beginning to be elucidated, leading to the suggestion that patient-specific
molecular profiling will provide information for the selection of treatment modality and dose-
intensity.6 Understanding of the signal transduction pathway of cellular radiation response has
paved the way for combining molecular and radiation therapy with very positive results in
clinical trials.7
The integration of advanced imaging techniques and knowledge of the molecular basis of cancer
will be a powerful influence on cancer management in general and radiotherapy planning in
particular. With this approach, one can perform radiobiological phenotyping to characterize the
tumor in terms of the factors that influence its radioresponse. A good example is the use of PET
for visualizing tumor hypoxia, which is known to be detrimental to the efficacy of radiation
treatment in several disease sites.8 In this case, the biological response of radiation could be re-
oxygenation that could be monitored by sequential PET imaging. Another promising assay of
biological response, based on an increased understanding of the mechanisms of cell death,
involves the use of radio-labeled Annexin V which adheres to the cellular membrane of apoptotic
cells.9 Lastly, the use of various PET tracers (with FDG being the current favorite) for
monitoring treatment response holds significant promise. 10 Methods based on magnetic
resonance are also being intensively evaluated as tools for planning and assessing radiation
treatment.11
It is clear that, given the complexity of biological systems, dose-based surrogates are inadequate
for describing the radiation responses of tumors and normal tissues. It is also certain that
molecular profiling and imaging will provide the approach for improving treatment design and
monitoring treatment response. What may be uncertain is the time course for attaining these
goals. In that regard, we physicists are very familiar with error and uncertainties estimates—but
whether it be a decade or two, the time is now to embrace the brave new world of biology-based
treatment planning.
Rebuttal
To cling to dose-based treatment planning, given the recent findings referenced in my opening
statement, is to keep "looking for the lost key under the light." We know that using purely
physical surrogates for biological endpoints is fundamentally wrong, and that dose was originally
12
chosen as the basis for treatment planning simply because at the time it was the only thing we
could quantify. But should one continue to hold to a system known to be fundamentally flawed
because the scientifically correct method is as yet imperfect? Shouldn't one start looking for the
key where it was lost by bringing light to that place?
Clearly, radiotherapy physicists must begin learning the pertinent aspects of tumor and radiation
biology to better understand the deficiencies of dose-based treatment planning vis-à-vis the
potentials (and pitfalls) of biology-based treatment planning. For example, one should know that
the assumption that "clonogenic-cell density and radiosensitivity distribution are constant
throughout the planning target volume" is wrong, and therefore the conclusion that "a uniform
dose distribution maximizes tumor control" of Dr. Li's third reference is incorrect as well.
Radiotherapy planning using biological information may or may not be perfected within the next
ten years, but even if not perfected it likely will surpass dose-based methods in predicting clinical
outcome. It's only in the last decade or so that analysis of dose-volume histograms has led to a
substantial increase in biological understanding of normal tissue complications, and that the
potentials of molecular profiling and imaging have emerged relative to the radioresponse of
tumors. The pace of these developments can only accelerate in providing the needed information
for optimizing radiation planning. It may be a decade or even two before the goals of treatment
based on biological end-points is realized, but the time is now for radiotherapy physicists to take
their heads out of the warm and fuzzy sand of dose-based treatment planning, and move towards
biologically-based treatment planning.
AGAINST THE PROPOSITION: X. Allen Li, Ph.D.
Opening Statement
The generation of treatment plans optimized to maximize tumor-cell killing and minimize
normal-tissue damage (biologically-based radiation treatment planning, BBRTP), rather than to
deliver a uniform dose to a specific treatment volume (physically-based RTP, PBRTP), has been
proposed for many decades. With the recent surge in intensity-modulated radiotherapy and
advances in biological/molecular imaging, interest in BBRTP is increasing within the
radiotherapy community. I believe that BBRTP will eventually replace PBRTP. However, the
central question of this Point/Counterpoint is "Will BBRTP be reliably practiced in the clinic
over the next decade?"
Despite many decades of effort and experience, knowledge of tumor and normal-tissue radiation
responses, on which BBRTP depends, is meager, unreliable, and not well documented. Unlike
PBRTP, biological models are essential to any BBRTP effort. Although there are several
phenomenologic and mechanistic models available, their predicting powers are generally poor
because of large uncertainties in the biological parameters. Efforts to develop such models have
increased in recent years, but progress has been slow, particularly for normal structures. Many
sources of uncertainty (e.g., treatment positioning, organ motion, dosimetry, the cell micro-
environment, irradiation history, and chemotherapy) contribute to current deficits in information
about normal tissue complications.12 In addition, unified tools to accurately document such
information remain to be developed. Although in-vitro data on radiation responses are abundant,
they can be applied in the clinic only with a great deal of caution.13
13
BBRTP requires patient-specific information on tumor/normal-tissue biology.

Biological/molecular imaging technologies, although they offer great promise in eventually
providing such information and even real-time monitoring of biological response, are far from
mature. Moreover, tumor and normal-tissue radiosensitivities most likely change during a course
of radiotherapy. The models, methods and technology needed to account for these phenomena
remain to be developed.
Because of uncertainties in tumor/normal-tissue biology, clinical use of BBRTP is risky for the
foreseeable future. For example, the use of incomplete or inaccurate biological information to
preferentially target selected regions of a tumor at the expense of other tumor regions may
jeopardize the treatment outcome. It may be possible, however, to use a combined
PBRTP/BBRTP approach. First, a treatment plan is generated using PBRTP. Then, BBRTP is
used to selectively target regions of the tumor that are potentially radioresistant without
decreasing the dose to any other region of the tumor. It is known that a uniform dose distribution
maximizes tumor control so long as the clonogenic-cell density and radiosensitivity distribution
are constant throughout the planning target volume. 14 If the combined plan can be accomplished
without appreciably increasing the dose to critical structures, the treatment outcome should be
improved. As we gain additional information about the efficacy of BBRTP, more aggressive
strategies may evolve. Because clinical trials require many years to complete, the replacement of
PBRTP with BBRTP as the standard of practice is certainly unlikely during the next decade. I
would like to acknowledge Dr. Robert Stewart for providing comments.
Rebuttal
As expected, Dr. Ling did a superb job explaining why dose-based surrogates are inadequate for
treatment planning. I agree with Dr. Ling that there is a compelling need for BBRTP. However,
the central issue we are debating is whether or not BBRTP will make a significant impact in the
clinic within the next decade. On this issue, our opinions diverge.
A major issue is that the models used to relate dose distributions to treatment outcomes are
highly nonlinear, and seemingly small perturbations in physical or biological factors can have a
large impact on estimates of biologic response, which is often organ/tumor and patient
specific.13,15,16,17 Dr. Ling suggests that advances related to the human genome project and
molecular imaging will provide the patient-specific information needed for BBRTP. There are
many hurdles that must be overcome before functional/molecular imaging will have a substantial
impact on the routine practice of radiation oncology, including imaging-pathology validation,
timing of post-therapy imaging, spatial and temporal evolution of tumors, and lack of clinical
outcome studies.18 For example, it is not clear whether PET or other imaging modalities can
accurately depict the microscopic extension of a tumor. The spatial and temporal variation of
hypoxia during radiation therapy poses a significant challenge for the use of PET or other
imaging modalities. Prospective multi-institutional clinical trials to evaluate the clinical impact
of biologically-based planning strategies and novel imaging techniques are almost nonexistent.
Until these issues are addressed, BBRTP will need to be applied in an incremental fashion. It
may be many years or even decades before BBRTP has demonstrated clinical benefits over
PBRTP. Nevertheless, I agree with Dr. Ling that we should embrace the brave new world of
BBRTP and actively pursue methods to speed the clinical implementation of these techniques.
REFERENCES
14
1. S. Levegrun, A. Jackson, M. J. Zelefsky, E. S. Venkatraman, M. W. Skwarchuk, W.

Schlegel, Z. Fuks, S. A. Leibel, and C. C. Ling, "Risk group dependence of dose response for
biopsy outcome after three-dimensional conformal radiation therapy of prostate cancer,"
Radiother. Oncol. 63, 11–26 (2002).
2. E. D. Yorke, A. Jackson, K. E. Rosenzweig, S. A. Merrick, D. Gabrys, E. S.
Venkatraman, C. M. Burman, S. A. Leibel, and C. C. Ling, "Dose-volume factors contributing to
the incidence of radiation pneumonitis in non-small lung cancer patients treated with three-
dimensional conformal radiation therapy," Int. J. Radiat. Oncol., Biol., Phys. 54, 329–339
(2002).
3. A. Jackson, M. W. Skwaarchuk, M. J. Zelefsky, D. M. Cowen, E. S. Venkatraman, S.
Levegrun, C. M. Burman, G. J. Kutcher, Z. Fuks, S. A. Liebel, and C. C. Ling, "Late rectal
bleeding after conformal radiotherapy of prostate cancer II: Volume effects and dose-volume
histograms," Int. J. Radiat. Oncol., Biol., Phys. 49, 685–698 (2001).
4. A. Niemierko, "Reporting and analyzing dose distributions: A concept of equivalent
uniform dose," Med. Phys. 24, 103–110 (1997).
5. R. Withers, J. Taylor, and B. Maciejewski, "Treatment volume and tissue tolerance," Int.
J. Radiat. Oncol., Biol., Phys. 14, 751–759 (1988).
6. C. N. Coleman, "Radiation oncology—linking technology and biology in the treatment of
cancer," Acta Oncol. 41, 6–13 (2001).
7. E. Nakata, N. Hunter, K. Mason, Z. Fan, K. K. Ang, and L. Milas, "C225 antiepidermal
growth factor receptor antibody enhances the efficacy of docetaxel chemoradiotherapy," Int. J.
Radiat. Oncol., Biol., Phys. 59, 1163–1173 (2004).
8. K. S. C. Chao et al., "A novel approach to overcome hypoxic tumor resistance: Cu-
ATSM-guided intensity-modulated radiation therapy," Int. J. Radiat. Oncol., Biol., Phys. 49,
1171–1182 (2001).
9. F. Blankenberg, C. Mari, and H. W. Strauss, "Imaging cell death in vivo," Q. J. Nucl.
Med. 47, 337–348 (2003).
10. C. Cohade and R. Wahl, "PET scanning and measuring the impact of treatment," Cancer
J. 8, 119–134 (2002).
11. K. L. Zakian, J. A. Koutcher, D. Ballon, H. Hricak, and C. C. Ling, "Developments in
nuclear magnetic resonance imaging and spectroscopy: application to oncology," Semin. Radiat.
Oncol. 11, 3–15 (2001).
12. T. E. Schultheiss, "The controversies and pitfalls in modeling normal tissue radiation
injury/damage," Semin. Radiat. Oncol. 11, 210–214 (2001).
13. D. J. Carlson, R. D. Stewart, X. A. Li, K. Jennings, J. Z. Wang, and M. Guerrero,
"Comparison of in vitro and in vivo  ratios for prostate cancer," Phys. Med. Biol. 49, 4477–
4491 (2004).
14. S. Webb, P. M. Evans, W. Swindell, and J. O. Deasy, "A proof that uniform dose gives
the greatest TCP for fixed integral dose in the planning target volume," Phys. Med. Biol. 39,
2091–2098 (1994).
15. J. Z. Wang, M. Guerrero, and X. A. Li, "How low is the  ratio for prostate cancer?"
Int. J. Radiat. Oncol., Biol., Phys. 55, 194–203 (2003).
16. M. Guerrero and X. A. Li, "Deriving radiobiological parameters from a large number of
clinical studies for radiotherapy treatment planning of early-stage breast cancer," Phys. Med.
Biol. 48, 3307–3326 (2003).
17. J. Z. Wang, X. A. Li, W. D'Souza, and R. D. Stewart., "Impacts of prolonged dose-
delivery time on tumor control: A note of caution for IMRT," Int. J. Radiat. Oncol., Biol., Phys.
57, 543–552 (2003).
18. S. Apisarnthanarax and K. S. C. Chao, "Current imaging paradigms in radiation
oncology," Radiat. Res. 163, 1–25 (2005).
15
1.4. Image-guided radiotherapy is being overvalued as a clinical tool

in radiation oncology
Howard I. Amols and David A. Jaffray
Reproduced from Medical Physics, Vol. 33, No. 10, pp. 3583–3586, October 2006
OVERVIEW
Advances in radiation therapy technology are coming fast and furious, with image-guided
radiotherapy (IGRT) being one of the latest developments to be offered by all linear accelerator
vendors. The radiotherapy community asked for this and now they have it. Without question,
knowing exactly where targets and normal tissues are at the time of each treatment is something
that all radiation oncologists would say they want. However, since IGRT systems are expensive
and time consuming to use, some would argue that, unless clinically significant improvements in
outcome can be demonstrated, the theoretical benefits of IGRT are not worth the added expense.
This is the premise debated in this month's Point/Counterpoint.
Arguing for the Proposition is Howard I. Amols, Ph.D. Dr. Amols received his Ph.D. in Nuclear
Physics from Brown University in 1974, followed by an NCI training fellowship in particle
therapy at Los Alamos National Laboratory. He has held medical physics positions at the
University of New Mexico, Brown, and Columbia Universities. He is currently Chief of Clinical
Physics at Memorial Sloan Kettering Cancer Center, where he sheepishly admits to spending a
good deal of his time working on IGRT. He has over 100 peer-reviewed papers, many of which
are on topics some consider to be even more overvalued than IGRT. He is certified by the ABMP
and is a Fellow, a Past President, and currently the Chairman of the Board of the AAPM. His
hobbies include horology and arguing just for the sake of argument.
Arguing against the Proposition is David A. Jaffray, Ph.D. Dr. Jaffray graduated in 1988 with a
B.Sc. in Physics from the University of Alberta and completed his Ph.D. in Medical Biophysics
from the University of Western Ontario in 1994. The following eight years were spent as a
Clinical Physicist in the Department of Radiation Oncology at the William Beaumont Hospital,
Royal Oak, MI. He is presently Head of Radiation Physics at the Princess Margaret Hospital and
Associate Professor in the Department of Radiation Oncology, University of Toronto, where he
holds the Fidani Chair in Radiation Oncology and Medical Biophysics. His principal research
interests involve the development of novel imaging systems for radiation therapy, including
amorphous-silicon-based large-area detectors and cone-beam CT, image-guided radiotherapy,
and integration of functional imaging into radiotherapy. Dr. Jaffray is certified in Radiation
Oncology Physics by the ABMP, currently serves as a member of the AAPM Science Council,
and is a frequent contributor to Medical Physics.
FOR THE PROPOSITION: Howard I. Amols, Ph.D.
Opening statement
That image-guided radiotherapy (IGRT) will improve radiotherapy is taken as axiomatic.

Improved patient setup accuracy reduces probability of geographic miss and improves treatment
16
outcome. But where are the data demonstrating that IGRT improves clinical outcome in a
clinically measurable way? If treatment margins (PTV minus CTV) are already larger than setup
uncertainties, and if such margins result in acceptable toxicity, what is gained with IGRT? Will
IGRT enable reduction of treatment margins? If so, will this ever be testable in a controlled
clinical trial? Given the uncertainties in guesstimating CTVs from GTVs, perhaps reducing field
margins (naively assuming that IGRT reduces errors) will be detrimental to treatment outcome?
Conversely, if one cannot reduce treatment margins what benefits are derived from IGRT?
Ling et al.1 have suggested that an IGRT system should have three-dimensional (3D) imaging of
soft tissues and tumors, plus an efficacious process for clinically meaningful intervention.
Currently IGRT fails on both counts. Image quality, degraded by the physics of kV cone-beam
scattering, is inferior to conventional CT; significant breathing artifacts result from the 30–60 sec
scan times. Marginal image quality coupled with imperfect and non-robust image registration
software both suggest questionable application of IGRT as currently practiced. Is using IGRT to
move a patient around daily by a few millimeters, in ignorance of resulting clinical outcome,
better than doing nothing?
Current IGRT systems also do a poor job of monitoring intrafractional errors. Unless these are
less than interfractional errors, using IGRT prior to but not during treatment accomplishes little.
Again, where are the data?
Even assuming that IGRT can detect setup errors, how should one correct for them? Is rigid-
body, bony anatomy registration of IGRT and CT-Sim images sufficient, or must one consider
soft tissue deformation? If the latter, then current IGRT systems are deficient. If the former, then
aren't existing MV electronic portal imaging systems adequate? When is kV and/or 3D required,
and when is two-dimensional (2D) MV sufficient?
Consider also the cost/benefit ratio: 3D-kV imaging systems cost $500,000, which is nearly 1/3
the cost of a linac. If one treats five patients per hour and IGRT adds three minutes this represents
a 25% increase. Maintenance, physics, and physician times also increase. Is IGRT worth the
price? Lamentably, its benefits will likely never be tested in a clinical trial as it is already widely
used without controls or forethought.
I'm not arguing that IGRT is a boondoggle. Indeed, it's counterintuitive to argue against
improved accuracy. But like many technological advances IGRT has come too fast, with too little
thought, and without serious discussion of which types of patients will benefit. American
hospitals have already collectively spent well over $100,000,000 on IGRT without a single peer-
reviewed randomized clinical trial demonstrating its benefits. This is another example of new
technology searching for a problem before the need has been demonstrated.
Lacking data, we've placed high expectations on an expensive, untested technology. As we

should have learned from Record and Verify systems, computerized patient setup doesn't
eliminate errors, it merely changes their nature.
I question not IGRT, but rather its current use. To quote Walt Kelly: "We have met the enemy and
they are us."
AGAINST THE PROPOSITION: David A. Jaffray, Ph.D.

17
Opening statement
Radiation therapy is a proven therapeutic agent directed at localized targets with the explicit
objectives of achieving control of the neoplasm while minimizing the risk of toxicity in
surrounding normal tissues. The relationship between geometric miss and failure in outcome is
one of the few relationships in radiotherapy that is not debated. The very fact that we are willing
to irradiate perfectly normal tissues to ensure tumor coverage highlights the importance of
succeeding in the accurate localization of dose and target.
Current practice of radiation therapy devotes substantial clinical and technical effort towards this
objective. The introduction of CT simulation, 3D treatment planning, portal imaging, and
verification of treatment parameters are existing infrastructure investments. Ongoing operating
costs are also substantial. These include daily manipulation of skin marks, checking for SSD
discrepancies, applying small shifts based upon portal imaging, and the time consumed in the
acquisition of port films and electronic portal images. We should also include time and effort
expended in the handling of exceptions by all of the disciplines involved in the management of
difficult or complex cases, so as not to have to withhold therapy for some patients due to our
inability to solve the technical challenges related to targeting. The recent introduction of IMRT
has heightened the concerns around geometric uncertainties as the opportunity to create highly
conformal dose distributions places additional pressure on executing the delivery with a known
geometric performance.2
While these advances have the best of intentions, it is remarkable to note how little quantitative
validation is performed for each patient to prove that a course of conformal therapy has been
executed as intended over the many fractions of radiation delivery. I would assert that there is a
growing inconsistency in the amount of effort placed on the design of therapy compared with the
assurance of its execution. In fact, the investments in further refinements in therapy that are
occurring upstream should be questioned compared with the clear and unanswered challenge of
placing radiation dose with geometric accuracy and precision within the human body. IMRT and
biological targeting could be criticized as "tails wagging the dog," however, we don't dictate the
order of innovation. IGRT technologies that provide volumetric imaging of both the targeted
structures and surrounding normal tissues provide the opportunity to move from an intention to
deliver state-of-the-art radiation therapy, to a patient-specific verification that the intention has
been satisfied.
There is a need to separate the issues of IGRT technology adoption from the more fundamental
underlying question: "What level of quality should be achieved in the delivery of radiation
therapy to the individual patient?" Simple population-based performance metrics ring hollow.
Clearly, PTV margins designed to cover the CTV in 95% of the patients would not go over well
with one patient in twenty. Geometric uncertainty in the localization of therapy is a major
controllable factor influencing the performance of radiation treatments, and it has been poorly
addressed for too long. Recent studies based upon portal imaging and repeat CTs demonstrate the
presence of patient subpopulations that are not well served by current positioning methods.3,4,5 It
is clear that the first step toward confirming the quality for each patient is to be able to measure
the geometric performance of the delivery. This requires imaging of internal structures at the
time of treatment, i.e., image-guided radiation therapy.
Rebuttal: Howard Amols, Ph.D.

18
There are three ingredients to good radiotherapy: correct tumor identification, conformal
treatment planning, and precise treatment delivery. With the advent of biological imaging (MR,
PET, etc.) we're only now learning how to do the first part correctly. We've very nearly perfected
the middle step, and as Dr. Jaffray points out, we're still relatively primitive with regard to the
third step. But his arguments that IGRT will improve that third step by a clinically significant, or
even measurable amount fall short of the mark. While I agree with his argument that IGRT has
the potential to improve the situation and should be tested, he makes no compelling argument to
justify hundreds of hospitals buying IGRT at this time to the tune of $100,000,000.
For example, he quotes his own studies,2,5 both of which use conventional CT rather than linac-
based cone-beam CT systems, simply to show that adaptive radiotherapy based on multiple CT
scans would allow delivery of slightly higher doses to prostate (on average, 7.5% more dose in
Ref. 2 and 13% in Ref. 5) without additional rectal toxicity. These are very idealized studies,
however, that assume no intrafraction motion and no organ deformation. They also ignore the
fact that linac-based IGRT systems have inferior image quality compared with conventional CT,
thus rendering them less accurate for identifying soft tissues, positioning patients, or monitoring
intrafraction motion. Furthermore, the prostate is a bad example to use to try to demonstrate the
utility of IGRT. Zelefsky et al.,6 for example, have already reported an eight-year actuarial PSA
relapse-free survival rate for favorable-risk prostate patients treated with IMRT to 81 Gy of
>85% with no grade 4 rectal toxicity and only 0.1% grade 3 toxicity, without the use of IGRT.
Thus, even if IGRT could buy you a 13% dose increment, it would likely be of little clinical
value. In fact, the benefits cannot be as large as Dr. Jaffray claims since he has ignored organ
deformation and intrafraction motion in his analysis. His other references (Refs. 3,4) are little
more than studies demonstrating that conventional 2D EPID improves the setup accuracy of
patients—not really news and not really a justification for a $500,000 3D IGRT system.
In diseases where radiotherapy does poorly, IGRT offers even less hope. In brain, for example,
IGRT offers little improvement over conventional stereotactic radiosurgery. In lung and upper
abdomen, IGRT will be almost useless unless it is respiratory gated, and current systems can't do
that. In head and neck, 2D EPIDs seem to work fine.
In short, Dr. Jaffray offers a hypothesis but no real clinical facts, and certainly no data justifying
IGRT's price tag or the enormous amount of hype currently associated with it.
Rebuttal: David Jaffray, Ph.D.
The advent of IGRT technologies has provided interesting data to awaken the community to, and
refocus our attention on, the targeting problem—something that we have neglected for too long.
The recent article by de Crevoisier et al.7 highlights the influence of systematic errors on
biochemical measures of outcome in radiation therapy of the prostate. This article describes a
retrospective study in which a correlation between rectal-filling-induced displacement of the
prostate at the time of planning is correlated with reduced biochemical control. I compliment the
authors for exploring this relationship and bringing their observations to the forefront. Daily
image-guidance data about the day-to-day positioning of the patient will elucidate the role that
important factors such as bowel filling,7,8 respiratory-induced motion,9 lung target deformation,10
and radiation-induced changes in the geometry of normal and target tissues11,12 have in limiting
the precision of radiation delivery.
At this point I have not advocated for margin reduction or dose escalation as a benefit of IGRT.
The arguments to date have related to quality assurance of current radiation therapy practice and
19
how the introduction of IGRT technologies allow us to recognize the elephant in the room
("geometrical miss") for what it is—a blind spot that limits outcomes and prevents understanding
of the effects of radiation therapy. The role of image guidance as a quality assurance and
education tool cannot be overstated.
So what about the future? Now that we have our dose in a pile, where does IGRT let us go? I
believe that the opportunities are enormous.
As there is no known benefit to unnecessary irradiation of normal tissues, IGRT will permit
margin reduction, hopefully applied with a strong understanding of the principles of image
guidance and residual uncertainties. The frequent argument used against PTV margin reduction is
that standard clinical practice is "hiding" disease in the PTV margins. This simply reflects the
need for more education on this topic—the reality is that we are "hiding" disease that may not be
there. The PTV is a geometric construct to assure dose to the clinical target volume, not a volume
known to contain any specific tissue. Clearly, we still have some work to do in educating the
community on these concepts.
One side effect of applying radiation therapy without confidence in dose placement is to maintain
a strict constraint on the degree of dose uniformity within the target. The emergence of
stereotactic radiotherapy protocols that permit hot spots provided "they are in the target" is a
harbinger of the importance of IGRT in moving away from the dogmatic and poorly supported
practice of delivering uniform dose to targets in radiation therapy. 13 Failure to walk away from
this practice within the next decade may truly limit innovation in radiation therapy. The
constraint of uniformity forces undue hardship on normal structures and will limit dose escalation
for those sites with poor control.
The radiation therapy community needs to remember that we are not standing by ourselves in the
field of oncology. The remarkable explosion in therapies grouped under the umbrella of
interventional radiology and targeted therapeutics is clearly exploiting advanced technologies
such as image guidance to maximize the benefit detected in outcome studies. 14 The accelerated
introduction of high-intensity focused ultrasound is made possible with the integration of MR-
guided placement and thermometry. It is my belief that radiation therapy wouldn't get out of the
FDA gate without image guidance if it were to be proposed anew today. Failure to advance our
skill in the application of radiation therapy will leave this proven therapy at the back of the pack
and will limit our ability to understand how this agent actually works in vivo.
Fundamentally, the adoption of IGRT is about quality in health care and our willingness to assure
high-quality delivery for every patient. I would argue that 100% of the patients should achieve
the best quality of radiation therapy if it is financially achievable. To bring this about we need to
drive industry to provide these tools such that they operate in the current cost envelope. Early
adoption is one way of achieving this objective. The answer to the question as to whether IGRT
is overvalued is obviously no. The real question is how we are going to innovate our practice to
make sure that it isn't overpriced.
REFERENCES
1. C. Ling, E. Yorke, and Z. Fuks, "From IMRT to IGRT: Frontierland or Neverland,"

Radiother. and Oncol. (in press).
20
2. A. A. Martinez et al., "Improvement in dose escalation using the process of adaptive

radiotherapy combined with three-dimensional conformal or intensity-modulated beams for
prostate cancer," Int. J. Radiat. Oncol., Biol., Phys. 50, 1226–1234 (2001).
3. L. E. Millender et al., "Daily electronic portal imaging for morbidly obese men
undergoing radiotherapy for localized prostate cancer," Int. J. Radiat. Oncol., Biol., Phys. 59, 6–
10 (2004).
4. M. H. Witmer et al., "Patient-specific daily pretreatment setup protocol using electronic
portal imaging for radiation therapy," J. Appl. Clin. Med. Phys. 6, 1–13 (2005).
5. M. Ghilezan et al., "Online image-guided intensity-modulated radiotherapy for prostate
cancer: How much improvement can we expect? A theoretical assessment of clinical benefits
and potential dose escalation by improving precision and accuracy of radiation delivery," Int. J.
6. M. J. Zelefsky et al., "Eight-year outcome of high-dose intensity modulated radiation
therapy for patients with clinically localized prostate cancer," Journal of Urology (in press).
7. R. de Crevoisier et al., "Increased risk of biochemical and local failure in patients with
distended rectum on the planning CT for prostate cancer radiotherapy," Int. J. Radiat. Oncol.,
Biol., Phys. 62, 965–973 (2005).
8. M. J. Ghilezan et al., "Prostate gland motion assessed with cine-magnetic resonance
imaging (cine-MRI)," Int. J. Radiat. Oncol., Biol., Phys. 62, 406–417 (2005).
9. C. Eccles et al., "Reproducibility of liver position using active breathing coordinator for
liver cancer radiotherapy," Int. J. Radiat. Oncol., Biol., Phys. 64, 751–759 (2006).
10. P. A. Kupelian, C. Ramsey, S. L. Meeks, T. R. Willoughby, A. Forbes, T. H. Wagner,
and K. M. Langen, "Serial megavoltage CT imaging during external beam radiotherapy for non-
small-cell lung cancer: observations on tumor regression during treatment," Int. J. Radiat.
Oncol., Biol., Phys. 63 (4), 1024–1028 (2005).
11. J. L. Barker, Jr., et al., "Quantification of volumetric and geometric changes occurring
during fractionated radiotherapy for head-and-neck cancer using an integrated CT/linear
accelerator system," Int. J. Radiat. Oncol., Biol., Phys. 59, 960–970 (2004).
12. M. B. Sharpe et al., "Adaptive planning and delivery to account for anatomical changes
induced by radiation therapy of head and neck cancer," Int. J. Radiat. Oncol., Biol., Phys., Suppl.
63, S3 (2005).
13. C. C. Ling and X. A. Li, "Over the next decade the success of radiation treatment
planning will be judged by the immediate biological response of tumor cells rather than by
surrogate measures such as dose maximization and uniformity," Med. Phys. 32, 2189–2192
(2005).
14. D. B. Zippel and M. Z. Papa, "The use of MR imaging guided focused ultrasound in
breast cancer patients; a preliminary phase one study and review," Breast Cancer 12(1), 32–38
(2005).
21
1.5. The value of PET/CT is being over-sold as a clinical tool in

radiation oncology
Lei Xing and Barry Wessels
Reproduced from Medical Physics, Vol. 32, No. 6, pp. 1457–1459, June 2005
OVERVIEW
PET and CT are important imaging technologies in the diagnosis and staging of cancer and in
monitoring and evaluating the effectiveness of treatment. Combined PET/CT units are being
marketed as advantageous because they offer a common platform for the technologies so that
patient convenience is improved and mis-registration of information is reduced. Some physicists
agree with this marketing message, while others believe that PET/CT is a technology that adds to
healthcare costs without contributing substantive value. This controversy is the subject of this
month's Point/Counterpoint.
Arguing for the Proposition is Lei Xing, Ph.D. Dr. Xing earned his Ph.D. in physics from the
Johns Hopkins University and obtained his medical physics training from the University of
Chicago. He is currently an Associate Professor of Radiation Oncology at Stanford University.
His major areas of interest are intensity modulated radiation therapy (IMRT), image-guided
radiation therapy (IGRT), PET/CT and other emerging molecular imaging modalities, such as
MR spectroscopic imaging (MRSI) and fluorescence/bioluminescent imaging, and their
integration with radiation therapy treatment planning systems. He has served on the AAPM
IMRT Subcommittee, and the Medical Physics Editorial Board. He is certified by the American
Board of Radiology.
Arguing against the Proposition is Barry Wessels, Ph.D. Dr. Wessels is currently Professor and
Director, Division of Medical Physics and Dosimetry at Case Western Reserve University. He
received his Ph.D. in nuclear physics from the University of Notre Dame in 1975. For the next
two decades, Dr. Wessels served as a group leader at the MIT reactor laboratory and as a faculty
member at The George Washington University. His areas of specialty include the dosimetry and
radiobiology of radionuclide therapy, image fusion and clinical trial design. Dr. Wessels is
currently a member of the MIRD and REIR committees of the SNM and has served as President
of the American Board of Sciences in Nuclear Medicine (ABSNM), Chair of the AAPM Task
Group on Radiolabeled Antibody Dosimetry, RTOG committee on physics and the ICRU
Nuclear Medicine Advisory Committee. He has published over 200 peer-reviewed journal
articles, book chapters, proceeding papers and abstracts.
FOR THE PROPOSITION: Lei Xing, Ph.D.
Opening statement
Hybrid PET/CT scanners1 are advertised as cutting-edge technology. Many clinics are making
purchasing decisions through the influence of advertisements and the pressure of competitors.
Since 2000, more than 300 PET/CT units have been purchased, even though the clinical benefits
22
of the technology have not been unequivocally demonstrated. At this point, the "hopes" for
PET/CT greatly outweigh the "facts" about the merit of PET/CT. Subjective impressions have
been the dominant factor influencing PET/CT purchase decisions.2,3
Hybrid PET/CT systems have several positive features that are absent with stand-alone PET and
CT units. PET/CT is a hardware-based image-fusion technology that virtually eliminates the
uncertainty and inconvenience of software fusion of separate PET and CT images that are
acquired with patients in different positions. However, PET images alone often contain enough
information to answer clinically-important metabolic questions3 without superposition of CT
images. For treatment planning in radiation therapy, the use of CT/PET fiducials and software
fusion has provided sub-centimeter spatial resolution4,5 and yielded excellent clinical results.6
Currently available fusion software performs well in the head and neck, and less well in the
thorax, abdomen and pelvis because of positioning difficulties and involuntary motion. However,
reasonable results can usually be acquired in these regions by use of PET transmission scans and
close collaboration of PET and CT technologists. Image co-registration is sometimes a problem
also with PET/CT in these regions, because the images are acquired sequentially rather than
simultaneously.7 Software fusion requires multiple steps and is tedious, but these problems will
ultimately be resolved through research.
The bottom line is that hardware-based fusion is not the ultimate solution for image co-
registration, and a hybrid machine is not needed for every combination of imaging technologies
that can conceivably yield useful fusion images. For example, a hybrid PET/MRI unit hardly
seems necessary, even though some researchers are considering it. It is indisputable that CT
measurements for PET attenuation corrections greatly reduces PET scan time—but it also ties up
the CT scanner and increases the radiation dose to the patient.
A hybrid PET/CT unit is not a "must-have" clinical tool in radiation oncology, and does not
automatically elevate an oncology service to "world-class" status. Conversely, an institution can
be "world-class" without a hybrid PET/CT unit.9 Software fusion, including rigid and deformable
image registration, has several virtues and should be targeted for further research. Further, the
possible clinical benefits of hybrid PET/CT should be carefully documented by investigators. At
the end of the day, the usefulness of any technology is justified only through definitive diagnostic
and therapeutic gains.
Rebuttal
Dr. Wessels puts his faith in marketplace trends, and bets that the marketplace cannot be very
wrong. That is exactly what concerns me. I caution that, historically, the marketplace can indeed
be wrong, and even badly wrong. If one is not convinced, perhaps revisiting the competition
between beta and VHS file formats as a video standard in the 1970s will help. That said, I do not
totally disagree with Dr. Wessels' argument, but would like to challenge him further by asking:
"Can the marketplace be that right?" The marketplace is an interesting arena to watch but should
never be a metric for technology assessment. Most likely, the truth of the debate about the value
of PET/CT lies somewhere between our respective Point/Counterpoint positions. Clinical and
scientific data are needed to objectively determine where PET/CT is located on the coordinate
starting from "totally wrong" to "absolutely right," whether the clinical benefit of PET/CT is
sufficient to offset its higher capital and running costs, and whether the alternative software
fusion is an option.
23
Dr. Wessels also mentions that the hybrid PET/CT facilitates clinical operation. While the hybrid
system is probably not a fundamental breakthrough that changes the paradigm of medical
imaging, it does provide a more convenient solution for radiation oncology practitioners
compared with the currently available software approach for fusion of PET and CT images. In an
era where time, convenience, and efficiency are money, this value should certainly not be
underappreciated. On the other hand, we should acknowledge that this benefit does not come for
free; there are a number of practical issues associated with it.3 Aside from the capital and
operational costs, increased radiation dose, changes in medical/administrative procedures,
appropriate selection of patients, etc., should all be considered. One should also note that
difficulties with current fusion software may be short-lived and things may improve in the near
future.
The fact sheet for the hybrid PET/CT is multi-dimensional. It is still too early to claim that
PET/CT is a technology for everyone, and that the application of PET/CT should not in any
circumstance be dictated by reimbursement. It is my view that if multi-modality imaging (not just
PET, but also other related modalities) is to have a genuine impact in radiation oncology, a
versatile, low cost, and long-term solution other than hardware fusion has to be in place.
AGAINST THE PROPOSITION: Barry Wessels, Ph.D.
Opening statement
Many spirited discussions occur within one or more of the following scenarios: 1) one discussant
is talking apples and the other oranges, 2) both positions have overlapping truth sets, and 3) the
definitive experiments have not been performed. Such is the case with the present "value"
discussion on the net worth of integrated PET/CT for radiation oncology. The fact that a New
England Journal of Medicine article in 2003 by Lardinois ,8 and a Special Supplement in the
January 2004 issue of The Journal of Nuclear Medicine, were devoted entirely to integrated
PET/CT (Ref. 9) suggests that something major is afoot. Proponents of the proliferation of
PET/CT machines would argue that the technology already has had a profound impact on the
restaging of disease and for radiation therapy treatment using expanded or contracted target
volumes for disease located in the head, neck, thorax and abdomen.10 The apples and oranges
aspect of this argument may be largely due to the simple addition of functional information to the
anatomical imaging process alone and may not have anything to do with the method of
acquisition. This addition can be accomplished by any one of three image registration methods
(visual side-by-side, software, and hardware fusion). The latter of these methods is currently
accomplished by the use of relatively costly PET/CT units. Clearly, the definitive experiments
have not been completed to resolve the argument over the value of PET/CT. Most studies
compare "horse-and-buggy" visual side-by-side fusion with inherently coregistered PET/CT
scans, and do not give relatively inexpensive, independent image fusion software packages an
even technical match.11 A common truth set for all applications is 1) the need for a reliable
automatic image segmentation algorithm for PET to draw corrected target volumes;12 2) the
requirement for machine-based fiducial alignment and patient immobilization systems;13 and 3)
respiratory gating for all.14 One might argue that this truth set is correct, but PET/CT may be the
wrong combination of machines. The pundits may have a point here; PET/MR is probably not far
around the corner.15
But at the end of the day from one who is constantly confronted with patient alignment, fusion,
throughput and convenience concerns, the integrated PET/CT scanner has a unique advantage to
24
potentially solve all of these problems in a single patient visit. The cost in physician, physicist,
radiation oncology and nuclear medicine staff time needed for multiple imaging sessions that
usually require redundant immobilization equipment in separate departments is just part of the
challenge of a nonintegrated PET/CT approach to image fusion. The expanded utility of PET/CT
for radiation therapy is driven not just by expert technocrats who have a vested interest in its
proliferation. It is also being demanded by referring physicians during combined tumor board
case presentations, and ultimately will be expected by patients. Is this not in the same class as the
unproven "value added" benefit purported by IMRT treatment for patients in radiation oncology?
To this end, can the marketplace be that wrong?
Rebuttal
Virtually all the "facts" stated by both my worthy opponent and me are in substantial agreement
regarding this discussion topic. It is heartening that most of the cited literature is shared common
ground. However, there is little agreement in the valuation of the cost/benefit ratio for both the
investigator and the patient concerning this new technological combination. My colleague states
that "software fusion requires multiple steps and is tedious" (agreed!), "but these problems will
ultimately be resolved through research." However, improvements to the software fusion process
may only add to the tedium with limited returns on effort. This problem can be generalized to all
image-guided therapy modalities, namely the functional definition of "real-time." To properly
gate for organ motion and provide reproducible patient positioning necessary for accurate
therapy, "real time" will be most likely defined in units of milliseconds. Using this metric, both
software fusion and PET/CT fail miserably. Typically for software fusion, "real-time" is on the
order of several hours to days for the acquisition of separate PET and CT scans taken in the
treatment position. For state-of-the-art PET/CT imaging using an attenuation map acquired from
CT, "real-time" is still fractions of an hour. The potential for patient movement, misalignment
and multiple organ motion remains problematic for both methodologies.
It is reasonable to hypothesize that the "real-time" motion problem will be substantially reduced
by the modality that acquires data over a window of time that is an order of magnitude closer to
the motion-stopping, millisecond goal. Presently, there is a growing library where software
fusion just gets it wrong for patients scanned in different positions using PET or CT originating
from separate devices, and where there is no hope for organ-motion gating. Performed by skilled
clinical staff, the single appointment PET/CT is the most practical solution to obtain the desired
data from a single machine for patients scanned in a predetermined therapy position.
REFERENCES
1. D. W. Townsend, J. P. Carney, J. T. Yap, and N. C. Hall, "PET/CT today and tomorrow,"

J. Nucl. Med. 45, 4S–14S (2004).
2. J. Czernin and H. Schelbert, "PET/CT imaging, facts, opinions, hopes, and questions," J.
Nucl. Med. 45, 1S–3S (2004).
3. W. V. Vogel, W. J. Oyen, J. O. Barentsz, J. H. Kaanders, and F. H. Corstens, "PET/CT:
Panacea, redundancy, or something in between?" J. Nucl. Med. 45, 15S–24S (2004).
4. S. Mutic, J. F. Dempsey, J. Markman, K. S. Chao, and J. A. Purdy, "Multimodality image
registration quality assurance for conformal three-dimensional treatment planning," Med. Dosim
26, 79–82 (2001).
5. J. N. Yu, F. Fahey, H. Gage, C. Eades, B. Harkness, C. Pelizzari, and J. Keyes,
"Intermodality, retrospective image registration in the thorax," J. Nucl. Med. 36, 2333–2338
(1995).
25
6. J. D. Bradley, F. Dehdashti, M. A. Mintun, R. Govindan, K. Trinkaus, B. Siegel, and J. S.

Welsh, "Positron emission tomography in limited-stage small-cell lung cancer: A prospective
study," J. Clin. Oncol. 22, 3248–3254 (2004).
7. B. Thorndyke, A. Koong, and L. Xing, "Reducing respiratory motion artifacts in
radionuclide imaging through retrospective stacking," Med. Phys.
8. D. Lardinois, W. Weder, T. F. Hany, E. M. Kamel, S. Korom, B. Seifert, G. K. Von
Schulthess, and H. C. Steinert, "Staging of non-small-cell lung cancer with integrated positron-
emission tomography and computed tomography," N. Engl. J. Med. 348, 2500–2507 (2003).
9. J. Czernin (Guest Editor), "PET/CT imaging function and structure," J. Nucl. Med. 45
(suppl), 1S–103S (2004).
10. I. F. Ciernik, E. Dizendorf, B. G. Baumert, B. Reiner, C. Burger, J. B. Davis, U. M.
Lutolf, H. C. Steinert, and G. K. Von Schulthess, "Radiation treatment planning with an
integrated positron emission and computer tomography (PET/CT): A feasibility study," Int. J.
11. P. J. Slomka, "Software approach to merging molecular with anatomic information," J.
Nucl. Med. 45 (suppl), 36S–45S (2004).
12. H. Schoder, Y. E. Erdi, K. Chao, M. Gonen, S. M. Larson, and H. W. D. Yeung, "Clinical
implications of different image reconstruction parameters for interpretation of whole-body PET
studies in cancer patients," J. Nucl. Med. 45, 559–566 (2004).
13. J. B. Bradley, C. A. Perez, F. Dehdashti, and B. Siegel, "Implementing biologic target
volumes in radiation treatment planning for non-small cell lung cancer," J. Nucl. Med. 45
(suppl), 96S–101S (2004).
14. S. A. Nehmeh, Y. E. Erdi, K. E. Rosenzweig, H. Schoder, S. M. Larson, O. D. Squire,
and J. L. Humm, "Reduction of respiratory motion artifacts in PET imaging of lung cancer by
respiratory correlated dynamic PET: Methodology and comparison with respiratory gated PET,"
J. Nucl. Med. 44, 1644–1648 (2003).
15. S. J. Nelson, M. R. Day, P. J. Buffone, L. L. Wald, T. F. Budinger, R. Hawkins, W. P.
Dillon, S. Huhn, M. D. Prados, S. Chang, and D. B. Bigneron, "Alignment of volume MR
images and high-resolution [18F]fluorodeoxyglucose PET images for the evaluation of patients
with brain tumors," J. Comput. Assist. Tomogr. 21, 183–191 (1997).
26
1.6. Respiratory gating for radiation therapy is not ready for prime
time
X. Allen Li and Paul J. Keall

Reproduced from Medical Physics, Vol. 34, No. 3, pp. 867–870, March 2007
OVERVIEW
Respiratory motion is a significant cause of geometric uncertainty for the radiation treatment of
lesions in the thorax region. This has to be taken into account in the design of planning target
volumes (PTVs), and this increases the volume of normal tissue that has to be irradiated. It is
well established that PTV size can be reduced if the effect of respiratory motion can be reduced.
There are many ways that this might be achieved such as specialized breathing control or breath-
hold methods, respiration-synchronized techniques, abdominal compression, and respiratory
gating. Since respiratory gating is the least intrusive and most patient friendly of these methods,
there have been numerous studies aimed at perfecting this technique and several respiratory
gating systems are now available commercially. This technology is, therefore, accessible to all,
but there is some concern that making this available in facilities that do not have the expertise to
realize its limitations and use it properly could be dangerous. Concern that release of such
methods on a wide scale at this time might be premature is the topic of this month's
Point/Counterpoint debate.
Arguing for the Proposition is X. Allen Li, Ph.D. Dr. Li is a Professor and the Chief of Medical
Physics in the Department of Radiation Oncology, Medical College of Wisconsin. He has been a
board certified medical physicist since 1994 and is currently a member of AAPM TG-74, TG-
106, and the Bio-effect Working Group. He is a member of the Editorial Board for Medical
Physics and a reviewer for seven scientific journals and four public and private research funding
agencies. He has authored 100 peer-reviewed papers and numerous abstracts. Dr. Li's research
interests range from Monte Carlo applications in radiation therapy to biologically based
treatment planning, and recently management of respiratory motion in radiotherapy.
Arguing against the Proposition is Paul J. Keall, Ph.D. Dr. Keall is Associate Professor and
Director of the Division of Radiation Physics in the Department of Radiation Oncology, Stanford
School of Medicine, CA. He obtained his M.S. in Health Physics and Ph.D. in Physics degrees
from the University of Adelaide, Australia. He serves the AAPM as a member of the Editorial
Board for Medical Physics and is a member of the Therapy Research Subcommittee, the
Stereotactic Body Radiotherapy TG102, the Monte Carlo in Treatment Planning TG105, and the
Summer Undergraduate Fellowship Program Subcommittee. His major research interests are 4-D
imaging, planning, and treatment, adaptive radiotherapy, respiratory gating, image-guided
radiotherapy, and biological models in radiotherapy.
FOR THE PROPOSITION: X. Allen Li, Ph.D.
Opening statement
27
Respiratory organ motion, which can be up to 30 mm, reduces the effectiveness of radiation
therapy (RT) for thoracic and abdominal tumor targets. This motion degrades anatomic position
reproducibility during imaging, demands larger margins during RT planning, and causes errors
during RT delivery. In recent years, a variety of methods and techniques have been proposed or
developed to explicitly account for respiratory motion in RT. These include respiratory gating,
breath holding, and respiration synchronization. The gating method, including internal and
external gating systems, is the most widely discussed method so far. It has been well documented
that gated RT, if carried out carefully, can significantly reduce margins and, thus, improve
sparing of normal tissues.1 Currently, the only internal gating system used in the clinic is
fluoroscopic tumor tracking based on implanted markers.2 For external gating technology, at
least two systems are commercially available for both CT and linear accelerators: the Siemens
Anzai pressure belt3 and the Varian RPM optical system.4 In my view, gated radiation delivery,
although promising, is not yet ready for prime time.
Generally speaking, respiration is an active and vital process not easily lending itself to
manipulation. The unpredictable variation in respiratory patterns during and between image
acquisition and day-to-day treatment delivery is the most challenging problem for all of the
techniques developed for managing respiratory motion. For a gated procedure, this variation can
result in inaccurate and/or inefficient dose delivery, and even geographic misses. Although much
effort has been expended in searching for effective methods to improve breathing reproducibility,
progress has not been remarkable. For example, one method to reduce breathing irregularity is to
train patients with audio prompting and/or visual feedback.5,6 It has been reported, however, that
about 50% of patients could not follow both audio and video instructions simultaneously.7 So far,
no reliable training method is available.
For external gating systems, the use of an abdominal motion signal as a surrogate for internal
tumor motion is not reliable because of variations in the correlation and phase shifts between the
surrogate and internal structure motion. 8,9 In addition, the currently available systems are limited
to one-dimensional motion. Although respiratory motion is predominantly in the superior-inferior
direction, there are exceptions.10 These problems can result in significant errors in dose delivery.
Furthermore, a robust treatment verification system capable of documenting the dose actually
delivered and preventing geographic misses is not available for gated delivery.
For internal gating systems, the high risk of pneumothorax due to the percutaneous insertion of
fiducial markers in the lung is a problem, as is also the high imaging dose required for
fluoroscopic tracking. These problems make internal gating impractical at the moment.
Respiratory motion is complex and patient specific, and it depends on location. It cannot be
predicted by any known medical/physiological models. 10 In addition, clinical parameters to
identify suitable patients for gating have not been defined. So far, the data acquired in the clinic
for gated radiation therapy have been mostly limited to demonstration of dosimetric benefits.
Whether these benefits transfer to outcome gain is unknown.
In conclusion, respiration gating, although very promising, is not mature and can be risky. Its
clinical benefits have not been documented. Therefore, for the moment, gated radiation therapy
should be performed with caution only for selected groups of patients.
AGAINST THE PROPOSITION: Paul J. Keall, Ph.D.

28
Opening statement
The recently published AAPM Task Group Report on respiratory motion management
recommended that ―If target motion is greater than 5 mm, a method of respiratory motion
management is available; and if the patient can tolerate the procedure, respiratory motion
management technology is appropriate.‖1 Respiratory gating is one technique for respiratory
motion management that has been seamlessly integrated into the clinical process. Does
respiratory gating solve all of the motion issues? Definitely not! But the debate is not ―Is
respiratory gating with current technology perfect?‖ The question is ―Is respiratory gating better
than not accounting for motion at all?‖ By disputing three fallacies I will argue that respiratory
gating is well established and widely available, and that there is strong clinical evidence that it
will result in improved patient outcomes.
Fallacy #1: Respiratory gating is not an established technology
The first published article on respiratory gating in radiotherapy was that of O'Hara et al.11 in 1989
in which they described seven patient treatments. Technologies that are now considered
mainstream but were actually first clinically used after respiratory gating include IMRT,
superposition and Monte Carlo-based treatment planning, multi-slice CT scanning, and cone
beam CT imaging. Respiratory gating is an established technology and it has been commercially
available in the US since 1999.
Fallacy #2: Not many sites have or use respiratory gating
Respiratory gating is the most widely available method of respiratory motion management with
several vendors offering respiratory gating products. Over 1000 respiratory gating systems have
been sold worldwide. In a recent RTOG IGRT survey, 26 out of 91 sites surveyed were
performing respiratory gating (J. Bradley M.D., ASTRO 2006), and the trend is toward increased
use. At Stanford, respiratory gating is performed on three linear accelerators with an additional
two performing respiratory tracking.
Fallacy #3: There are no clinical data to support the use of gating
The proof is in the pudding as the saying goes—is respiratory gating better for patients? A
seminal paper by Fang et al.12 demonstrated that 3D radiotherapy has statistically significant
survival advantage over 2D radiotherapy, with 27% (3D) vs. 6% (2D) overall 5-year survival. A
compelling piece of data in that study was that patients treated with respiratory gating had a
hazard ratio of 0.25 (p=0.05), indicating that patients are four times more likely to survive than
those not treated with gating. In multivariate analysis gating was not significant—the study was
not powered to address this question—but there is a strong suggestion that respiratory gating
does, indeed, improve survival in lung cancer patients. Wagman et al.13 in a liver cancer study
found that using respiratory gating reduced fluoroscopically visible motion from
22.7 to 5.1 mm. The use of gating, in combination with a rigorous portal imaging protocol,
allowed the CTV-PTV margin to be reduced from 2 to 1 cm. This margin reduction allowed
dose increases from 7% to 27%, and also allowed radiotherapy to be given to patients for whom
the treatment would otherwise have been too toxic. Several treatment planning studies have
reinforced these clinical findings.14,15,16,17,18,19
Perhaps the most accurate form of respiratory gating has been implemented by the Hokkaido
group,20,21 where implanted markers are tracked in real-time using fluoroscopy, and the treatment
29
beam gated on when the markers are within predetermined positions. This active program has
treated many liver and lung cancer patients with gated stereotactic radiotherapy.
In summary, if respiratory motion management systems are available, and have shown clinical
benefit, we are ethically bound to use them. Today at Stanford we are routinely treating the
following abdominal and thoracic sites with respiratory gating and IMRT/conformal
radiotherapy: lung, breast, esophagus, pancreas, and lymphoma. Respiratory gating is beyond
prime time, it is routine.
Rebuttal: X. Allen Li, Ph.D.
Respiratory gating, which associates with tight margins, is better than not accounting for motion
at all only if it is carried out properly. Tight margins, which increase the risk of geometric
misses, require accurate treatment delivery. However, such accurate delivery is not trivial during
a course of multi-fractionated radiotherapy because of inter- and intrafractional variations. I
agree with Dr. Keall that the technology for respiratory gating is reasonably established and the
technical implementation of commercial gating systems can be seamless. However, the clinical
use of these technologies is not yet mature. The fact that only 26 out of 91 RTOG members are
performing respiratory gating reflects the hesitation of using these technologies for patients on a
large scale. Gating technology has been commercially available since 1999 and most RTOG
members are academic centers that would be expected to have the resources needed to perform
respiratory gating, but only a few are actually using it. Issues that contribute to this hesitation
include (1) more effective prediction techniques and/or training methods for patient breathing are
needed to reduce inter- and intrafractional variations in breathing patterns; (2) more reliable
mechanisms are required to correlate external surrogates with internal tumor motions; and (3)
proper clinical parameters are needed to identify suitable patients. Until these issues have been
fully addressed, respiratory gating should be carried out only in those clinical settings that can
provide the considerable resources needed to carefully select suitable patients, cautiously design
treatment plans, and extensively validate treatment delivery.
Rebuttal: Paul J. Keall, Ph.D.
I concur with Dr. Li's statements that, if used appropriately in conjunction with methods to
manage interfraction variations, respiratory gating allows safe margin reduction. I also agree with
the statement that respiratory motion is complex, multidimensional, and patient specific.
However, it is precisely for these reasons that we want to manage motion with technologies such
as respiratory gating. Ignoring the problem does not make it go away—respiratory gating can
reduce the apparent motion by over a factor of 4 (Ref. 13) and can minimize deleterious effects.
Following are my responses to the statements of Dr. Li:
High imaging dose for internal gating: The dose for fluoroscopy-based gating has been estimated
to be 1% of the treatment dose.20
Unacceptable risk of pneumothorax for marker implantation: Percutaneous22 and

bronchoscopic23 lung implantation have shown pneumothorax rates of 19% and 2%, respectively.
The increased accuracy offered by marker implantation outweighs the associated morbidity.
50% success rate of audio-visual biofeedback: A 24-patient repeat session study found a 100%
success rate, reducing the residual motion by an average of 20% at exhale and 25% at inhale.5
30
Lack of verification: The same verification methods used without gating can be used with gating,
e.g., portal images. However, as respiratory motion affects many of the landmarks used for planar
imaging, such as the diaphragm, chest wall, and carina, verification is even more effective with
respiratory gating than without.
To conclude, respiratory gating is at its prime time now. It represents an intermediate step
towards the widespread implementation of target tracking technology.
REFERENCES
1. P. J. Keall, G. S. Mageras, J. M. Balter, R. S. Emery, K. M. Forster, S. B. Jiang, J. M.

Kapatoes, D. A. Low, M. J. Murphy, B. R. Murray, C. R. Ramsey, M. B. Van Herk, S. S.
Vedam, J. W. Wong, and E. Yorke, ―The management of respiratory motion in radiation
oncology report of AAPM Task Group 76,‖ Med. Phys. 33, 3874–3900 (2006).
2. H. Shirato, S. Shimizu, T. Kunieda, K. Kitamura, M. van Herk, K. Kagei, T. Nishioka, S.
Hashimoto, K. Fujita, H. Aoyama, K. Tsuchiya, K. Kudo, and K. Miyasaka, ―Physical aspects of
a real-time tumor-tracking system for gated radiotherapy,‖ Int. J. Radiat. Oncol., Biol., Phys. 48,
1187–1195 (2000).
3. X. A. Li, C. Stepaniak, and E. Gore, ―Technical and dosimetric aspects of respiratory
gating using a pressure-sensor motion tracking system,‖ Med. Phys. 33, 145–154 (2006).
4. H. D. Kubo, P. M. Len, S. Minohara, and H. Mostafavi, ―Breathing-synchronized
radiotherapy program at the University of California Davis Cancer Center,‖ Med. Phys. 27, 346–
353 (2000).
5. R. George, T. D. Chung, S. S. Vedam, V. Ramakrishnan, R. Mohan, E. Weiss, and P. J.
Keall, ―Audio-visual biofeedback for respiratory-gated radiotherapy: impact of audio instruction
and audio-visual biofeedback on respiratory-gated radiotherapy,‖ Int. J. Radiat. Oncol., Biol.,
Phys. 65, 924–933 (2006).
6. V. R. Kini, S. S. Vedam, P. J. Keall, S. Patil, C. Chen, and R. Mohan, ―Patient training in
respiratory-gated radiotherapy,‖ Med. Dosim. 28, 7–11 (2003).
7. T. Neicu, R. Berbeco, J. Wolfgang, and S. B. Jiang, ―Synchronized moving aperture
radiation therapy (SMART): improvement of breathing pattern reproducibility using respiratory
coaching,‖ Phys. Med. Biol. 51, 617–636 (2006).
8. N. Koch, H. H. Liu, G. Starkschall, M. Jacobson, K. Forster, Z. Liao, R. Komaki, and C.
W. Stevens, ―Evaluation of internal lung motion for respiratory-gated radiotherapy using MRI:
Part I--correlating internal lung motion with skin fiducial motion,‖ Int. J. Radiat. Oncol., Biol.,
Phys. 60, 1459–1472 (2004).
9. D. P. Gierga, J. Brewer, G. C. Sharp, M. Betke, C. G. Willett, and G. T. Chen, ―The
correlation between internal and external markers for abdominal tumors: implications for
respiratory gating,‖ Int. J. Radiat. Oncol., Biol., Phys. 61, 1551–1558 (2005).
10. Y. Seppenwoolde, H. Shirato, K. Kitamura, S. Shimizu, M. van Herk, J. V. Lebesque,
and K. Miyasaka, ―Precise and real-time measurement of 3D tumor motion in lung due to
breathing and heartbeat, measured during radiotherapy,‖ Int. J. Radiat. Oncol., Biol., Phys. 53,
822–834 (2002).
11. K. Ohara, T. Okumura, M. Akisada, T. Inada, T. Mori, H. Yokota, and M. J. Calaguas,
―Irradiation synchronized with respiration gate,‖ Int. J. Radiat. Oncol., Biol., Phys. 17(4), 853–
857 (1989).
12. L. C. Fang, R. Komaki, P. Allen, T. Guerrero, R. Mohan, and J. D. Cox, ―Comparison of
outcomes for patients with medically inoperable Stage I non-small-cell lung cancer treated with
two-dimensional vs. three-dimensional radiotherapy,‖ Int. J. Radiat. Oncol., Biol., Phys. 66(1),
108–116 (2006).
31
13. R. Wagman, E. Yorke, P. Giraud, E. Ford, K. Sidhu, G. Mageras, B. Minsky, and K.

Rosenzweig, ―Reproducibility of organ position with respiratory gating for liver tumors: use in
dose-escalation,‖ Int. J. Radiat. Oncol., Biol., Phys. 55, 659–668 (2003).
14. L. E. Butler, K. M. Forster, C. W. Stevens, C. Bloch, H. H. Liu, S. L. Tucker, R. Komaki,
Z. Liao, and G. Starkschall, ―Dosimetric benefits of respiratory gating: a preliminary study,‖ J.
Appl. Clin. Med. Phys. 5(1), 16–24 (2004).
15. G. Starkschall, K. M. Forster, K. Kitamura, A. Cardenas, S. L. Tucker, and C. W.
Stevens, ―Correlation of gross tumor volume excursion with potential benefits of respiratory
gating,‖ Int. J. Radiat. Oncol., Biol., Phys. 60(4), 1291–1297 (2004).
16. M. Engelsman, G. C. Sharp, T. Bortfeld, R. Onimaru, and H. Shirato, ―How much margin
reduction is possible through gating or breath hold?‖ Phys. Med. Biol. 50(3), 477–490 (2005).
17. R. W. Underberg, J. R. van Sornsen de Koste, F. J. Lagerwaard, A. Vincent, B. J.
Slotman, and S. Senan, ―A dosimetric analysis of respiration-gated radiotherapy in patients with
stage III lung cancer,‖ Radiother. Oncol. 1, 8 (2006).
18. R. W. Underberg, F. J. Lagerwaard, B. J. Slotman, J. P. Cuijpers, and S. Senan, ―Benefit
of respiration-gated stereotactic radiotherapy for stage I lung cancer: an analysis of 4DCT
datasets,‖ Int. J. Radiat. Oncol., Biol., Phys. 62(2), 554–560 (2005).
19. S. S. Korreman, A. N. Pedersen, M. Josipovic, L. R. Aarup, T. Juhler-Nottrup, L. Specht,
and H. Nystrom, ―Cardiac and pulmonary complication probabilities for breast cancer patients
after routine end-inspiration gated radiotherapy,‖ Radiother. Oncol. 80(2), 257–262 (2006).
20. H. Shirato, S. Shimizu, K. Kitamura, T. Nishioka, K. Kagei, S. Hashimoto, H. Aoyama,
T. Kunieda, N. Shinohara, H. Dosaka-Akita, and K. Miyasaka, ―Four-dimensional treatment
planning and fluoroscopic real-time tumor tracking radiotherapy for moving tumor,‖ Int. J.
Radiat. Oncol., Biol., Phys. 48(2), 435–442 (2000).
21. H. Shirato, S. Shimizu, T. Kunieda, K. Kitamura, M. van Herk, K. Kagei, T. Nishioka, S.
Hashimoto, K. Fujita, H. Aoyama, K. Tsuchiya, K. Kudo, and K. Miyasaka, ―Physical aspects of
a real-time tumor-tracking system for gated radiotherapy,‖ Int. J. Radiat. Oncol., Biol., Phys.
48(4), 1187–1195 (2000).
22. Q. T. Le, B. W. Loo, A. Ho, C. Cotrutz, A. C. Koong, H. Wakelee, S. T. Kee, D.
Constantinescu, R. I. Whyte, and J. Donington, ―Results of a phase I dose escalation study using
single-fraction stereotactic radiotherapy for lung tumors,‖ J. Thor. Oncol. 1(8), 802–809 (2006).
23. M. Imura, K. Yamazaki, H. Shirato, R. Onimaru, M. Fujino, S. Shimizu, T. Harada, S.
Ogura, H. Dosaka-Akita, K. Miyasaka, and M. Nishimura, ―Insertion and fixation of fiducial
markers for setup and tracking of lung tumors in radiotherapy,‖ Int. J. Radiat. Oncol., Biol.,
Phys. 63(5), 1442–1447 (2005).
32
1.7. Heterogeneity corrections should be used in treatment planning

for lung cancer
Nikos Papanikolaou and Eric E. Klein

Reproduced from Medical Physics, Vol. 27, No. 8, pp. 1702–1704, August 2000
OVERVIEW
Corrections in dose distributions for the heterogeneous nature of lung tissue are available with
many commercial treatment planning platforms of recent vintage. Use of these corrections is
appealing to medical physicists who strive for accuracy in the display of radiation dose in
exposed tissues. However, radiation oncology is largely an empirical science. Treatment
regimens for patients, including those with lung cancer, are determined largely by experience.
This experience has been acquired without the luxury of heterogeneity
corrections. Inserting heterogeneity corrections into the planning process can be construed as
switching horses in midstream. In addition, some physicists question whether the heterogeneity
corrections have been verified sufficiently for all potential applications. These issues are
explored in this issue of Point/Counterpoint.
Arguing for the Proposition is Nikos Papanikolaou, Ph.D. Dr. Papanikolaou is currently on the
faculty at Emory University School of Medicine. He received master and doctorate degrees in
Medical Physics from the University of Wisconsin-Madison. He has lectured nationally and
internationally on dose computation
algorithms and heterogeneity corrections for photon beams. He is currently the chair of the
AAPM Task Group on Inhomogeneity Corrections. Dr. Papanikolaou is certified by the
American Board of Radiology and is actively involved with the American Association of
Physicists in Medicine, American College of Medical Physics, and European Society of
Therapeutic Radiation Oncology.
Arguing against the Proposition is Eric Klein, M.S. Mr. Klein received his master degree in
Applied Physics from the University of Massachusetts in 1987. He has been working as a
Radiation Oncology Physicist since 1981, the last 11 years at the Mallinckrodt Institute of
Radiology, where he is currently an Assistant Professor. Mr. Klein is on the editorial boards of
Medical Physics and the International Journal of Radiation Oncology, Biology, and Physics. He
was Scientific Co-Director for the 1999 AAPM Annual Meeting, and is a Track Chair for the
2000 World Congress Meeting, and Scientific Director for the 2001 AAPM Annual Meeting. Mr.
Klein is a member of AAPM‘s TG-65 on Tissue Inhomogeneity Corrections for Photon Beams.
FOR THE PROPOSITION: Nikos Papanikolaou Ph.D.
Opening Statement
To do or not to do? That is the question. There are indeed a number of issues to consider when
discussing the inclusion of heterogeneity (inhomogeneity) corrections in lung dose computations.
33
Is the dose calculation algorithm accurate that is used for the computation of heterogeneity
corrections? Several algorithms have been proposed to implement some sort of inhomogeneity
correction, from the simplified ratio of TARs (RTAR) that yields a correction factor for water-
based calculations, to convolution and Monte Carlo methods that include the inhomogeneity in
the calculation of patient dose. The latter, and specifically the convolution/superposition
algorithm that is becoming the standard for 3D radiation treatment planning systems, can
accurately predict the dose even in areas of electronic disequilibrium. Monte Carlo can yield
improvements over convolution computations, especially when inhomogeneities of nontissue
atomic numbers are involved. Monte Carlo is of lesser importance for lung irradiation. When
dose is computed two questions should be answered: (i) does the algorithm account for the three
dimensional shape of the patient, including inhomogeneities and, if so, (ii) does it do so for both
the primary and the scatter energy transport? The best algorithms meet both of these
requirements, and correctly predict the absolute dose to the prescription point and the dose
distributions that reflect the spatial distribution of inhomogeneities. The latter is particularly
important in the case of lung radiotherapy, since the shape of the isodose curves that encompass
the tumor volume will change significantly when lung inhomogeneity computations are used.
Similarly, the dose to the isocenter for a given number of monitor units (MU) changes (typically
increases) because of increased transmission through the lung.
If we are using an algorithm that correctly transports primary and scatter energy, we have to
consider the integration of that information in the prescription dose. However, the medical
physicist is not responsible for deciding the prescription dose! Nevertheless, it is our
responsibility to educate our physician colleagues and help them interpret treatments when
inhomogeneity corrections are included. It has been argued that tissue inhomogeneity corrections
should not be used at all since current clinical practice and dose prescriptions
are largely based on historical experience with tissue response to doses computed in water.
However, clinical data indicate that even for simple anterior–posterior parallel-opposed fields
encompassing the mediastinum and lung regions, corrections at the reference target point can
range between 0.95 and 1.16, and may be even larger for oblique or lateral fields. Because the
entire dose distribution can be distorted by tissue inhomogeneities, not only is the dose to the
target volume affected, but also the doses to nearby critical
organs which often impose limits to the prescription dose.
On a final note, we know that the level of dose accuracy required for clinical trials depends on
the ability to demonstrate a statistically significant improvement in clinical results as the dose is
altered. A side but important benefit of including inhomogeneity corrections is the impact on the
number of patients required in clinical trials. Inhomogeneity corrections reduce the uncertainty in
absolute dose, yielding more controlled studies with less variability in absolute dose delivery.
This feature facilitates treatment outcome studies and interinstitutional comparisons.
Clearly, standardization of absolute dose delivery cannot be achieved if one allows uncontrolled
variability caused by the anatomy, geometry, and density associated with individual patients.
Accurate doses, including inhomogeneity corrections, are an essential component of dose
optimization and the objective analysis of clinical results, especially with the advent of 3D
precision conformal radiotherapy.
Rebuttal
It is precisely because lung cancer radiotherapy is a complex matter that we have to move away
from the current ‗‗canned‘‘ practice of treatment delivery. How can we improve cure rates for
34
lung cancer unless we deliver higher doses to the tumor in a precise manner? We can‘t escalate
the dose if we are uncertain about the actual dose to the healthy lung, heart or esophagus, or to
the tumor itself. The findings of RTOG 88-08 support this argument: a 33% spread in delivered
doses was reported when inhomogeneities were included in calculations. The report concluded
that: ‗‗Existing density-correction algorithms are accurate enough to significantly reduce these
variations.‘‘ All of the pieces of the puzzle are here: (i) we now have more accurate algorithms
for dose calculation (convolution/superposition) than we had before; (ii) workstations provide
full 3D computations in only a few minutes; (iii) imaging technology has improved so that we
can obtain better, more accurate images faster, even allowing us to gate the images to reduce
motion; (iv) accelerator design has improved so that we can deliver (MLC and dynamic MLC)
and monitor (portal imaging) treatments more efficiently, including gated treatment options. So
why wait? Let‘s move forward now and do it right!
AGAINST THE PROPOSITION: Eric E. Klein, M.S.
Opening Statement
The use of heterogeneity corrections in treatment planning for lung cancer radiotherapy
exemplifies the cliche´ ‗‗the more you look, the more you find.‘‘ The desirability of
heterogeneity corrections is still evolving and unresolved. The physics community is gearing the
radiotherapy community to a condition where isodoses, treatment time calculations, and most
importantly, prescriptions account for the presence of lung. But until algorithms evolve to
accommodate this condition, prescriptions should not change. Heterogeneity corrections are not
like the recent TG-43/NIST SK change applied to I-125 and Pd-103, where prescriptions
changed to deliver the same prescribed dose as the clinician specified before the change. The
influence of lung tissue on dose delivery is more complex because of patient variations,
variability of algorithm sophistication, and resulting biological considerations. The radiotherapy
community is still refining HDR prescriptions, and has just begun to debate intensity modulated
radiotherapy prescriptions.
Lung cancer radiotherapy involves complex organs at risk, including healthy lung tissue, liver,
heart, esophagus, spinal cord, and brachial plexus. Dose distributions to these organs are
obviously affected by low-density lung tissue and also beam energy. To further complicate the
issue is the variation of a patient‘s lung density depending on the patient‘s status and
inspiration/expiration capabilities. Volume and density can differ, day to day, breath to breath,
and, more importantly, from scan to treatment. These issues may be resolved with improved
imaging information and respiratory control systems. This brings us to calculations.
The AAPM Science Council‘s Radiation Therapy Committee first assigned a task group to
investigate this issue in 1985. To date, no report has been generated, indicating the complexity of
the issue. From the simplest correction techniques of linear attenuation (ratio of TARs) and
Batho, to today‘s convolution (FFT and superposition) and Monte Carlo techniques, correction
methods have provided the best corrections possible at their period of technological development
provided that the corrections are implemented and utilized properly, and their limitations are well
understood. Accordingly, national clinical protocols have wisely examined the effect of
corrections in a postplanned manner, where corrections are applied using water-based monitor
units and uncorrected beam weights. The RTOG 88-08 lung radiotherapy protocol asked
facilities to submit water-based plans and post-corrected plans, along with information on the
planning algorithm used. Patients were planned, prescribed and treated without corrections. The
35
post-dosimetric analysis was a comprehensive effort during a 2D planning era, and was therefore
limited compared with volumetric dose calculation systems that are now available. This brings us
to RTOG 93-11, a 3D dose escalation lung radiotherapy protocol. This protocol also requires
prescription and treatment without corrections, and with the unit density and post corrected dose
distributions submitted. Early submissions included data using corrections that fall short of what
is commercially available today. In other words, clinical experiences and dosimetric analyses
never coincide with the most up-to-date technologies. Further accrual of data in response to
RTOG 93-11 should bring us closer to knowing the dose delivered after applying corrections,
and therefore how prescriptions would be changed to account for the corrections. The timing of
modern day algorithms and prescriptions may then finally be coincident.
Rebuttal
Dr. Papanikolaou and I agree that heterogeneity corrections should be used in treatment planning
for lung cancer. We differ on the timing of when this should take place. Although I concur that
the needed algorithm sophistication is in place, there are still unsettled issues. First, there is no
way to mandate or determine that a physicist understands and has properly parameterized, tested,
and implemented a sophisticated algorithm. This concern includes the question of how a
physicist analyzes dose distributions and calculates monitor units. For example, the algorithm
may accurately depict a dose reduction at the periphery of a lung tumor, while the computation
of dose based on a reference point located in the middle of the GTV calls for a reduction of
monitor units. This contradictory situation could lead to underdosage of the tumor periphery, if
one bases monitor unit calculations on an increased transmission through the lung.
I also concur with Dr. Papanikolaou that using proper algorithms and applying resulting
prescriptive corrections will reduce the uncertainty in dose reporting for patients enrolled in
national protocols. But I do not agree that responsibility for prescriptions is exclusively a
physician‘s obligation. We cannot dictate change unless we can offer advice on how to change.
So how do we offer advice? Once the results of heterogeneity corrections on dose distributions
have been ascertained from RTOG 93-11, we could consider a plan to change prescriptions. This
plan must also include the effects of lung density corrections on critical organ doses.
In closing, the most important task a physicist has in treatment planning is to accurately report
the dose to every element of the irradiated volume, targeted or not. But we should not let
clinicians make prescriptive changes haphazardly without advising them properly. We can and
will soon be able to do this.
36
1.8. The rate of evolution of radiation therapy planning and delivery

systems is exceeding the evolution rate of quality assurance
processes
Sheri D. Henderson and Peter J. Biggs

OVERVIEW
The processes of planning and delivering radiation treatments are experiencing a significant
upscaling in sophistication and complexity. Computer-based treatment planning systems utilize
medical images, radiation source characterization data, and desired treatment parameters to
determine preferred treatment options and predicted dose distributions. Treatment options
include low and high dose rate techniques (brachytherapy) and conformal, intensity-modulated
and image guided methods (external beam therapy). Medical physicists are responsible for
ensuring that the planning and delivery of radiation treatments meet agreed-upon standards for
quality assurance. These standards reflect a consensus of medical physicists that is difficult to
achieve when techniques are rapidly changing. Whether this dilemma currently exists in
radiation therapy planning and delivery is the subject of this Point/Counterpoint issue.
Arguing for the Proposition is Sheri D. Henderson. Dr. Henderson has worked as an independent
consultant for the last 10 years in areas of clinical physics, QA of equipment, commissioning and
data con figuration of treatment delivery and planning systems; FDA regulatory standards for
product introduction, specification, hazard analysis, hardware and software analysis, and testing
and documentation of radiation algorithm performance. Prior to becoming a consultant, Sheri
had an academic career as a research scientist for the clinical charged particle program at
Lawrence Berkeley Laboratory and as an Assistant Professor of Therapeutic Physics at
Mallinckrodt Institute of Radiology. Sheri received her Ph.D. in Radiation Biophysics from the
University of Kansas.
Arguing against the Proposition is Peter J. Biggs. Dr. Biggs obtained his Ph.D. in high energy
physics from the University of London in 1966. He worked at the Daresbury Nuclear Physics
Laboratory until 1970 and at MIT from 1970 until 1975, performing research at Fermilab and
Brookhaven National Laboratory, resulting in the discovery of the J particle. Since 1975, he has
been at the Massachusetts General Hospital in the Department of Radiation Oncology where he
is an Associate Professor and currently acting head of Radiation Biophysics. He has been active
on many AAPM task groups, including, most recently, TG51, the new AAPM dosimetry
protocol and TG57, an update of NCRP#49 for radiation therapy; he is currently chairman of
PUBCOM.
For the proposition: Sheri D. Henderson
Opening Statement
The processes of radiotherapy planning (RTP) and delivery are experiencing a significant
upscaling in enhanced capability and complexity. Past upward movements in these technologies
37
have not always occurred in concert over time, thereby contributing to the dilemma debated here.
Currently, the medical physicist is challenged to ensure that innovations in RTP and delivery
systems meet agreed-upon standards for quality assurance (QA). These standards, however, are
generally created much later than initial market introduction of the technologies.
At the beginning of this decade, introduction of the dynamic or virtual wedge in radiation
delivery preceded RTP implementation and medical physics QA. Although introduced into the
market with representative dose data, dynamic and virtual wedges were only sparingly used in
the clinics for several years before RTP systems started implementing segmented treatment
tables (STTs) for dose calculations, and linear arrays became available for verification. A similar
situation existed for introduction of independent collimation and multi-leaf collimators, except
these features enjoyed clinical use almost immediately while RTP systems are improving
implementation today. Although many scientific papers cover these topics with respect to dose,
absent is an agreed-upon QA standard for physicists.
In the middle of this decade, ‗‗3D‘‘ RTP systems with associated graphics and dose calculation
algorithms were introduced. This event challenged the medical physicist because QA required a
higher understanding of the RTP algorithms. The fact that a symposium is being held in 1999 for
basic Monitor Unit calculations for photons reflects the lack of a national AAPM-recommended
formalism in the presence of more complex treatments utilizing new technologies. Another
indication is a 1999 symposium exploring the use of DOSEGEL for 3D dosimetry. Even more
disturbing, however, is the willingness of medical physicists to purchase ‗‗canned‘‘ treatment
delivery data loaded into installed RTP systems.
Today, one only has to read inquiries on the medphys server or review current educational
symposia offered by the medical physics community and industry to realize that we are
endeavoring to master the sophistication (real and imaginary) of RTP systems in conjunction
with the technological advances of delivery systems. This state is worrisome as the delivery
systems are emerging with technologies of intensity modulation and inverse planning. Certainly,
there is doubt among medical physicists that for these new technologies, physics understanding
exists, RTP algorithms are sophisticated enough, and QA standards are established to ensure
accurate delivery of dose to patients.
Rebuttal
Upon reading Dr. Biggs‘ position statement, it is apparent that we disagree on our views of
history with regard to introduction of new technologies and the subsequent response of the
medical physics community to provide adequate QA safeguards. I suspect that our different
perspectives may reflect our different experiences. However, I believe that my experience has
more viewpoints than Dr. Biggs‘ because I have existed in the ‗‗ivory tower‘‘ of academia,
worked with industry, and labored in the (often isolated) trenches of most practicing medical
physicists.
I agree with Dr. Biggs that manufacturers adequately test products prior to introduction.
Manufacturers are required by law to practice GMP (good manufacturing practices) and meet
product specifications of performance. I also agree that it is the physicist‘s job to bring the
product into clinical operation. However, I disagree with Dr. Biggs‘ sweeping statement that (all)
medical physicists do this with extensive QA procedures that have been disseminated widely in
the literature and by professional organizations. My experience in the ‗‗field‘‘ just doesn‘t
support this claim.
38
Complex technologies are being installed at sites that previously performed only simple hand
calculations and Cobalt therapy. They do not have an on-site physicist or QA equipment, and
data are often taken by non-qualified individuals. In a recent seminar, a dozen competent
physicists could not agree on necessary QA components, the need for each, and the approach to
calculation of monitor units generated for a 3D complex treatment plan even after referring to the
standard text by Geoff Ibbott and Bill Hendee.
I do not believe that our differing view of ‗‗medical physics history‘‘ or today‘s debate will be
resolved without a measuring bar of standards. The issue of agreed-upon QA standards is not
addressed by Dr. Biggs, whereas it is foremost in importance from my view ‗‗from the road.‘‘
Against the proposition: Peter J. Biggs
Opening Statement
The rationale behind this proposition is that we are at the threshold of a computer-based
technological revolution that is likely to have a far larger impact on radiation therapy than any
experienced hitherto. The fear is that with so many possibilities for new treatment techniques,
such as Intensity Modulated Radiation Therapy (IMRT), coupled with increasingly complex 3D
planning systems, insufficient effort will be directed toward adequate quality assurance.
However, the history of medical physics practice in radiation therapy and the innate
conservatism of medical physicists make that scenario highly unlikely. Medical physicists do not
embark on new procedures without extensive testing to ensure that calculations, measurements
and plans make physical sense. When manufacturers produce either new hardware, such as
multileaf collimator (MLCs), or develop new software, such as dynamic wedge or dynamic
MLC, these products are tested for efficacy and safety. However, they are not commissioned by
the manufacturer—it is the task of individual physicists to bring them into clinical operation.
When electron beams first became available, the primary concern of the manufacturers was to
ensure that the beams were flat and had a sharp penumbra. It was left to the physics community
to perform the necessary dosimetry before patient treatments began. A comparison between
commissioning an MLC compared with a dynamic wedge illustrates physicists‘ conservatism;
that for the MLC is relatively short and straightforward, accounting for its rapid and universal
acceptance throughout the community, whereas for the dynamic wedge, it is more complicated
and thus less well accepted.
New-found knowledge is rapidly disseminated throughout the medical physics community by a

network that includes the Medical Physics journal, the AAPM annual and local chapter meetings,
summer schools, AAPM task groups, personal contacts and now e-mail listservers.
Is the current technological change far greater than hitherto experienced? Complex technology
has been introduced in recent years with adequate quality assurance safeguards. Stereotactic
radiosurgery comes readily to mind. This procedure requires positioning accuracy to better than
1 mm for treating lesions a few cm in diameter. This could not have been achieved without
additional mechanical devices to assist in the set-up and QA protocols carried out before each
procedure. Moreover, 3D treatment planning has been steadily developing over many years; the
principles of 3D planning at the Massachusetts General Hospital using beam‘s eye view and dose
volume histograms began with proton therapy in the mid 70‘s, almost 25 years ago!
Interestingly, many of the dose calculation algorithms developed for the first generation
computers are still in use today. QA instrumentation has also kept pace with the technology of
39
dose delivery. For example, detector arrays are now available for dynamic wedge
characterization and hardware/software is available for daily, monthly and annual machine
checks.
While errors have occurred in adoption of computer-controlled technology, such as Malfunction

54, to the best of this author‘s knowledge, no other similar faults have occurred since then. This
is a bold testimony to the adequate QA procedures that both manufacturers and medical
physicists have instituted to avoid repeating such mistakes.
The first sentence of a thesis abstract published in the March edition of Medical Physics provides
a fitting summary: ‗‗This thesis deals with the implementation of IMRT into the clinic and a
method of quality assurance which can be used on each intensity-modulated beam prior to
treatment.‘‘
Rebuttal
Dr. Henderson points out the apparent contradiction that while advanced technologies, such as
IMRT, are being introduced into the clinic, at the same time a symposium on basic monitor unit
(MU) calculations is being held. While IMRT is practiced at several centers, these are major
teaching hospitals that have invested significant amounts of time in planning and, more
importantly, verifying the dose distributions in the time-honored way of measuring them with
film or ion chambers. The need for a symposium on basic MU calculations does not imply that
some physicists are calculating correctly and others not, only that there are several methods to
achieve the same end. A symposium on DOSEGEL is entirely appropriate at this time since it is
a maturing technology and may have a great impact on the field of 3D QA. Dr. Henderson‘s
comments about the delay in the clinical introduction of dynamic wedges echo those I made and
are a healthy sign that physicists adopt a careful approach.
A high percentage of academic and community centers now use CT as the primary method of
simulation with direct input into a treatment planning computer. I believe this favorable adoption
is because the planning can vary from sophisticated 2D to full-blown 3D; the physicists can
therefore begin with familiar methodologies and progress to more advanced planning at their
own speed. These new technologies also bring new benefits: more accurate customization of
blocks through beam‘s eye view, dose-volume histograms and digitally reconstructed
radiographs for comparison with verification simulations and treatment portals.
40
1.9. Long-term changes in irradiated tissues are due principally to

vascular damage in the tissues
John Hopewell and H. Rodney Withers
Reproduced from Medical Physics, Vol. 25, Issue 12, pp. 2265-2268, December, 1998
OVERVIEW
Changes in irradiated tissues that occur months or years after exposure are one of the potential
complications of radiation therapy. The biological processes underlying these changes are
disputed among radiation biologists. Some experts believe that the changes are a reflection of
radiation-induced damage to the vascular system supplying the tissues, whereas others suggest
that the radiation-induced damage is to the tissues themselves. These two perspectives on a
fundamental radiobiological problem are discussed in this Point/Counterpoint series.
Arguing for the Proposition is John Hopewell. Professor John Hopewell obtained his Ph.D. from
the University of London in 1968 for work related to the effects of ionizing radiation and
chemical carcinogens on the central nervous system. He moved to the University of Oxford in
1970 and was appointed Director of Radiobiological Research at the University of Oxford in
1980. The Research Group he heads has focused its attention on the study of the mechanisms of
late radiation responses following therapeutic or accidental radiation exposure. Additional
research has assisted in the development of the present ICRP Radiological Guidelines for the
skin and has helped to improve our understanding of the problems of dose, time, fractionation
and volume effects in radiotherapy.
Arguing against the Proposition is H. Rodney Withers. Dr. H. Rodney Withers was born in
Queensland, Australia. He received his medical degree from the University of Queensland, and
his Ph.D. and D.Sc. degrees from the University of London. His research interests have been in
radiation biology and how it relates to radiation therapy for cancer. He has worked at the Gray
Laboratory in England, the National Cancer Institute, the M.D. Anderson Cancer Center, the
Prince of Wales Hospital in Sydney, and at UCLA where he is Professor and Chair of the
Department of Radiation Oncology and holds an American Cancer Society Clinical Research
Professorship.
For the proposition: Professor John Hopewell
Opening Statement
Early pathological reports on tissues, showing late radiation-induced changes after therapeutic or
accidental exposure, almost unanimously emphasized the prominent appearance of gross
vascular changes. These changes were varied in their appearance, affecting arteries/arterioles,
capillaries and veins. While an association was frequently suggested it was perhaps Rubin and
Casarett (1968), in their classical textbook ‗‗Clinical Radiation Pathology,‘‘ who first proposed a
causal relationship between the slow development of vascular damage and the occurrence of late
normal tissue injury. Their proposal initiated a spectrum of studies specifically designed to
examine long term time- and dose-related changes in the vasculature of complex tissues and in
41
simpler models. Such models included the vessels of the mesentery in the mouse, the hamster
cheek pouch and even the choroid plexus of the rat brain. The simpler models allowed easier cell
and vessel identification. The work was largely driven by the desire to demonstrate target cell
populations, a reflection perhaps of an extrapolation from the same approach in acutely
responding normal tissues.
These studies demonstrated, in general terms, that both endothelial cells and smooth muscle cells
were gradually lost from the wall of blood vessels after irradiation. The magnitude of the loss
was dose-related. Endothelial cell loss occurred earlier than that of smooth muscle cells, a feature
that could be interpreted by cell proliferation studies to indicate a more rapid turnover of
endothelial cells. Evidence for atypical endothelial regeneration has been reported resulting in
the appearance of groups of cells (colonies) at irregularly spaced intervals along the wall of a
blood vessel. Blood vessels with focal occlusion changes could often be visualized in vivo.
Suggestions that these groups of cells represented clonogenic cell survival were again driven by
classical radiobiological concepts.
Histological studies have indicated that there are occasions where intimal proliferation is so
severe that vascular occlusion may result. Associated studies have reported a concurrent
reduction in the vascular density of tissues, the severity of which was dose-related. These
changes, not surprisingly, were frequently linked to a reduction in regional blood flow. The
concept of a reduction in vascular density, and impaired blood flow leading to the development
of ischemic necrosis, is best identified by studies on pig skin. Pig
dermal tissue is relatively poorly vascularized compared with rodent dermal tissue, and this has
been used to explain the difference in dermal radiation response in the two species. Similarly in
the central nervous system, the presence of white matter necrosis, in the absence of similar grey
matter effects, has been explained by their differing vascular architecture. Whether simple
vascular insufficiency and ischemia is the explanation for white matter necrosis in the CNS
remains an open question.
More recent studies have focused on functional changes that may develop in endothelial cells
either as a direct consequence of irradiation or, more likely, as a result of a reduction in
endothelial cell number. The changes are numerous and include, (i) an upregulation of adhesion
molecules on the surface of endothelial cells, with the resultant adherence and infiltration of
white blood cells into tissues and (ii) modified eicosanoid metabolism, specifically an imbalance
in the two key eicosanoids, prostacyclin and thromboxane A2. White blood cell infiltration has
been clearly identified to be a factor in the development of white matter necrosis in the CNS and
in the development of glomerular sclerosis.
Pharmacological intervention after irradiation, based admittedly on insufficiently understood

vascular mediated effects, has also been effective in reducing the incidence and severity of late
effects. Successful outcomes of this type have been reported in the skin, CNS, kidney and lung.
Recent studies in the rat spinal cord have perhaps best exemplified the role of endothelial
damage, which sets up a cascade of events leading to the development of overt late tissue
damage. Irradiation was with thermal neutrons alone or in combination with a 10B capture agent
(borocaptate sodium—BSH). This agent does not cross the intact blood brain barrier. The
liberation of short range fission products (<9 m) from the 10B capture reaction, 10B(n, )7Li,
allows preferential irradiation of vessels and a relative sparing of the parenchyma. Nevertheless,
white matter necrosis was identical irrespective of the mode of irradiation. Studies of glial
progenitor survival at iso-effective doses for late damage showed marked differences in both
42
short and longer term responses, indicating their lack of involvement in the development of
necrosis. Furthermore, irradiation involving heterogeneous exposure with sparing of the
parenchyma, shortened the latency for white matter necrosis compared with that following
uniform tissue irradiation with thermal neutrons. It was implied from these results that irradiation
of the parenchyma may actually impair the development of the cascade of events initiated by
endothelial vascular injury. On this basis I strongly support the proposition that late effects
of radiation are due principally to vascular damage.
Rebuttal
Tissues with a hierarchical structure are frequently thought of as acutely responding tissues, i.e.,
skin, oral mucosa, etc. The timing of their response will be dependent on the specific total
turnover time of that tissue; hence oral mucosal reactions will occur earlier than those in pig or
human skin. The way that such tissues react to injury is to increase proliferation in the stem cell
compartment. The lens of the eye is also an epithelial tissue, but with a slower turnover, hence
cataracts appear at late times via a mechanism that is well understood.
Dr. Wither‘s main contention is that major variations in the rate of development and
radiosensitivity of late effect response tissues and organs are not consistent with a single target
structure, blood vessels. By this he implies that blood vessels are rather uniform structures.
However, vascular organization varies greatly both between tissues and organs and even within
an individual organ. Some vascular networks are adapted for a specific function, i.e., the
capillary network of a glomerulus, with a single input and a single output from the capillary
network. Other capillary networks have extensive collateral links. Unlike epithelial tissues
increased stem cell proliferation is not the usual initial means of response of the vasculature to
injury; physiological compensation is possible although variable from tissue to tissue. For
example blood flow in normal skeletal muscle can increase tenfold even under moderate
exercise.
We have both made reference to the central nervous system (CNS); white matter, or myelinated
areas, are verypoorly vascularized compared with grey matter regions. Its vascular physiological
reserve is also minimal compared with that found, and indeed needed, for the demands of the
grey matter regions of the CNS. Moreover, even in the myelinated regions, which are the most
sensitive to the development of necrosis, variations exist. The fimbria of the hippocampus
responds more rapidly to radiation exposure than the adjacent corpus callosum. This fact has
been related to a difference in vascular supply between these two white matter regions of the
brain. The fimbria has what is a terminal vascular supply, while the corpus callosum received a
more extensive collateral supply from adjacent grey matter regions making it less vulnerable to
the effects of vascular impairment. Differentiated oligodendrocytes, and more specifically their
progenitors, have recently been eliminated as a potential target cell population for radiation
induced white matter necrosis.
This the diversity of vascular systems could well explain many of the differing patterns of
response observed.
Against the proposition: H. Rodney Withers
Opening Statement
43
Contrary to the general belief that acute effects of irradiation result from killing of parenchymal
cells and late effects from vascular injury, I contend that both types of effect result directly from
radiation induced parenchymal cell depletion. Vascular damage occurs just as in other late
responding tissues. While vascular damage may exacerbate other types of the late injury it is not
the primary cause.
The rate of development of overt injury depends upon the rate at which cells of the tissue divide:
acute injury occurs in rapidly proliferating tissues, late injuries in tissues which turn over slowly.
A late effect is analogous to an acute effect but delayed in its expression by the slow rate at
which the target cells turn over. That late radiation injury develops at different rates in different
tissues is surely not consistent with a single target structure ~blood vessels!. Also, endothelial
injury appears appreciably earlier than parenchymal injury in most late responding tissues. Even
if late responses were attributed to non-endothelial elements of the vasculature, a consistently
different response rate from tissue to tissue should not occur. The most plausible explanation for
the variability in latent intervals lies in the variability of parenchymal proliferation patterns in the
various tissues which manifest late responses.
Not only do the rates of development of late effects vary widely among tissues, but so do the
doses required for a certain level of injury. Approximate tolerance doses in 2 Gy fractions are
kidneys 20 Gy, lungs 15 Gy, heart 40 Gy, brain 50 Gy, bladder 65–70 Gy, and dermis 70 Gy.
These differences could be easily explained in terms of varying initial numbers of ‗‗target‘‘ cells
in different tissues and/or variations in their radiosensitivity, whereas it is difficult to implicate
blood vessel damage as a universal basis for effects requiring such a wide range of doses.
Another interesting fact is that the RBE n/ varies widely among late responding tissues whereas if
all late effects were due to vascular injury it should be constant.
Then there is the nature of the lesions themselves. In spinal cord or brain, the characteristic
change is demyelination, implying a loss of glial cells, specifically oligodendrocytes. Why would
vascular damage target the myelin producing cells and not the neurons? More likely, the slowly
proliferating oligodendrocytes die slowly and the non-proliferating neurons survive, while the
blood vessels do their own independent thing. And radiation injury does not develop in
sympathetic nerves or ganglia, nor in the posterior pituitary, nor primarily in the gray matter of
spinal cord or brain. What makes these non-responsive structures different from the other parts
of the nervous system which are susceptible to radiation injury is not that they do not have a
vascular supply but that they do not have myelination. Why should myelin-making cells, which
turn over slowly, die a vascular death when all other cells are allowed an apoptotic or mitotic
death?
One late effect which cannot be the result of vascular damage is cataract formation because the
lens has no blood vessels.
Vascular damage happens: blood vessels do not have immunity from radiation. There are a
variety of changes in blood vessels to which a role in the etiology of a universe of late effects has
been inappropriately assigned. Regardless of what structural or functional injury to blood vessels
is chosen as the culprit, vascular degeneration can compromise organ function, but such
compromise is more sizzle than steak: the real target cells are specific to the organ, and why
should they be considered immune to radiation injury?
Rebuttal
44
Dr. Hopewell describes many different vascular changes after irradiation but does not explain
how they lead to a wide spectrum of late effects. One vascular effect he lists, intimal hyperplasia,
could be a direct result of irradiation which causes parenchymal atrophy, or it could be caused by
radiation-induced parenchymal atrophy, in which case it would be called degenerative, or
atrophic endarteritis: the chicken and egg story.
The response of the central nervous system to alpha particles produced from boron neutron
capture (BNC) is not a compelling argument. First, why does similar damage to vessels in the
grey matter not produce injury there? Second, alpha particles may pound the vessel walls but
they also penetrate the white matter. Third, for spinal cord, the RBE for high LET radiations
versus x rays is high (reflecting a low alpha/beta ratio for x rays) and so BNC alpha-particles
should produce glial injury at relatively low doses, and quickly.
John Hopewell and I have disagreed on this topic for decades. I still hope that a late effect of this
discussion will be his conversion to my point of view.
45
1.10. The best radiotherapy for the treatment of prostate cancer

involves hypofractionation
John F. Fowler and Alan E. Nahum
Reproduced from Medical Physics, Vol. 33, Issue 9, pp. 3081-3084, September, 2006
OVERVIEW
In a recent publication, Fowler et al.1 proposed that hypofractionation (fewer fractions and higher
dose/fraction) should be better than conventional fractionation for the treatment of prostate
cancer because of the low for prostate cancer reported in several studies. A low  value
means that prostate cancer cells have an unusually high capacity for repair of sublethal damage at
low doses and low dose/fraction. Hence, Fowler et al. argue, low dose/fraction should be avoided
and patients should be treated with hypofractionation. However, Nahum et al.2 have suggested
that the low  values reported for prostate cancer are artifacts caused by ignoring the effect of
radioresistant hypoxic cells present in about 20% of prostate cancers. They claim that the true
 for prostate cancer is about the same as for other types of cancer and therefore
hypofractionation should be avoided. We have brought two outstanding scientists to the table to
debate this important issue.
Arguing for the Proposition is John F. Fowler, D.Sc., Ph.D. Professor Fowler is Emeritus
Professor of Human Oncology and of Medical Physics at the University of Wisconsin, Madison.
He retired in 2004 and is now living back in London, England. Dr. Fowler's degrees include a
Ph.D. in Radiation Physics in 1956 and a D.Sc. in Radiation Biology in 1974 from the University
of London, as well as an Honorary MD in 1981 from the University of Helsinki. Dr. Fowler
started his career as a medical physicist, and was Professor of Medical Physics at the Royal
Postgraduate Medical School, Hammersmith, London, from 1963–1970. He then turned his
talents to experimental radiobiology and was appointed Director of the Gray Laboratory in
London, retiring from there in 1988. He then moved to the University of Wisconsin, Madison,
where he spent a total of 11 years (1988–1994 and 1999–2004) as a bio-mathematical modeler in
radiation oncology. Dr. Fowler's extensive bibliography includes over 500 research papers,
numerous book chapters, and a book on heavy particle radiotherapy.
Arguing against the Proposition is Alan E. Nahum, Ph.D. Dr. Nahum's Ph.D. is on Theoretical
Radiation Dosimetry from Edinburgh University in 1975. After four years teaching science in
schools he joined the Medical Physics Department at Umeå University in Sweden, where his
work was primarily on ion-chamber correction factors and dosimetry codes of practice, including
a sabbatical at NRCC, Ottawa with Dave Rogers. From 1985 to 2002 he worked at the Royal
Marsden Hospital/Institute of Cancer Research, UK, where his research dealt mainly with
"biological modelling," especially tumor control probability (TCP) and normal tissue
complication probability (NTCP). Short spells at Fox-Chase Cancer Center, Reggio Emilia, and
Copenhagen followed. Dr. Nahum currently works at the Clatterbridge Centre for Oncology in
the UK and is Visiting Professor at Liverpool University, where his major interest is in using
TCP and NTCP models to optimize radiotherapy treatment plans.
46
FOR THE PROPOSITION: John F. Fowler, D.Sc., Ph.D.
Opening Statement
Four clinical analyses of the ratio of prostate tumors were published between 1999 and 2003
yielding values of 1.5, 1.45, 1.2 and 3.1 Gy.3,4,5,6,7 The last differed from the first three in that
unrealistically early repopulation start-up times (0–28 days) were assumed, otherwise the 
value would have been similar to the others. Provided the  ratio for prostate tumors is not
above that for late rectal damage (about 3 Gy), hypofractionation should be safe. Using a
calculated reduced total dose (and overall times not too short), there should be no change in late
complications or tumor control from conventional schedules. Hypofractionation to reduce
fraction number is clearly the "best treatment" for patients' convenience and for less costly
healthcare.
Hypofractionation is safe, especially when delivered with intensity modulated radiotherapy

(IMRT). In London, 47–55 Gy in 12 fractions and 36 Gy in 6 fractions were used for prostate
radiotherapy without major complications. 8 The Manchester group recently reported that 705
patients treated with 16 fractions of 3.1 Gy achieved equivalence to 33 fractions of 2 Gy,9 which
is only true if  is about 1.5 Gy. It is clear that 28 fractions of 2.5 Gy are feasible,10 as are 22
fractions of 2.94 Gy (Ritter, personal communication), 20 fractions of 3 Gy (Toronto and UK,
personal communications), 9 fractions of 6 Gy,11 5 fractions of 6.7 Gy,12 6 fractions of 6.9 Gy,13
and 4 fractions of 10 Gy.13
The intriguing aspect is that there may be a biological bonus for patients. If  is significantly
lower than 3 Gy then, for equal late rectal reactions, the biological effect on the tumor should be
greater for hypofractionated than for conventional treatments. Surprisingly large gains are
predicted if =3 and 1.5 Gy for rectum and prostate tumor, respectively. 1 For a schedule of 10
fractions of 4.9 Gy, the rectal normal tissue tolerance dose (in 2-Gy equivalent fractions) is only
78 Gy for a tumor dose equivalent of 91 Gy.1
Evidence is mounting that  is indeed about 1.5 Gy. Lukka et al.14 randomized 936 patients to
20 fractions of 2.62 Gy vs 33 fractions of 2 Gy in a Canadian trial. The ratio was determined
to be 1.12 Gy (95% confidence interval (CI ): –3.3 Gy, 5.6 Gy,).15 This is one of the best datasets
and illustrates the problem of identifying  values within clinical 95% limits. For example, an
Italian nonrandomized trial16 with 334 patients yielded a point value of = 9.8 Gy (95% CI:
0.7 Gy, 16 Gy,)15 but, because of the wide confidence interval, this abnormal  value can be
discounted as evidence that the for prostate cancer is high.
From the work of Demanes et al.,17 we know that  is less than 3.0 Gy. They published 5 to 10
year results of 209 patients treated with external beam radiotherapy (20 fractions of 1.8 Gy) +
high dose rate (HDR) brachytherapy (4 fractions of 5.8 Gy). For intermediate-risk prostate they
found 96% tumor control at 5 years, yielding 1.2 Gy (0.05 Gy, 1.9 Gy, 95% CI). Modeling
using =3 Gy would predict 75% tumor control (chi-squared difference from 96% = 10.3, p =
0.001).18 This demonstrates that  is significantly less than 3 Gy, and opens the door to
expecting tumor gains with hypofractionation.
47
I claim the debate. Further clinical trials should be done to improve precision of ; but we need
fear neither the loss of prostate tumor control from hypofractionation, nor complications, if dose
reduction is done appropriately.
AGAINST THE PROPOSITION: Alan E. Nahum, Ph.D.
Opening Statement
Current conventional external-beam radiotherapy for prostate tumors involves between 64 and
90 Gy delivered in 2-Gy fractions. The higher doses are made possible by employment of 3D
conformal therapy, most recently IMRT, at many radiotherapy clinics. Local control rates are
generally excellent except for advanced disease. 19 Additionally, impressive control rates have
been obtained for early-stage disease using low dose rate brachytherapy with I-125 seeds.20 Thus,
in general, radiotherapy for prostate tumors works well when delivered using conformal
techniques.
Where, then, does the idea of hypofractionation come from? The answer is that it comes from
radiobiological modeling in which the (biochemical) control rates from external-beam therapy
have been compared to those from low dose rate brachytherapy.4 Several such studies5,7,21 have
claimed to demonstrate that the  value for prostate cancer is about 1.5 Gy, which is lower than
that for the surrounding normal tissues. This led my jousting partner Professor Fowler to
advocate fraction sizes much larger than 2 Gy, coupled with a corresponding decrease in the total
dose to ensure isoeffective complication rates.1 Yes, if it really is true that the effective  for
prostate clonogens is of the order of 1.5 Gy, then this hypo-strategy should yield higher control
rates than the current 2 Gy/fraction treatment schedules. But is the for prostate cancer low?
Are there any published modeling analyses which do not yield a low  Yes.
Firstly, the use of a tumor control probability (TCP) model incorporating both inter-patient
variation in radiosensitivity and hypoxia,2,22 together with mean radiosensitivity values ()
taken from the radiobiological literature, suggested that  need not be low. In fact, the mean
 value derived from in-vitro radiobiological clonogenic assays was 8.3 Gy.2
Secondly, can we deduce anything from documented treatment outcomes involving relatively
large fractions? One thinks of the Christie Hospital in Manchester which has routinely employed
~3-Gy fractions. In fact, their recently published prostate outcome data 9 are consistent with the
predictions of the heterogeneous radiosensitivity-hypoxia TCP model referred to above,2,23 i.e.,
this also did not suggest that prostate  is low.
Thirdly, if a low  favors hypofractionation then conversely it must impede a

hyperfractionated schedule. Recently, however, a group in Milan reported on a 1.2 Gy twice-a-
day vs 2 Gy once-a-day treatment series.16 Were control rates worse for the patients treated with
1.2 Gy fractions? Not at all. Outcome was markedly better for the 1.2 Gy than the 2 Gy patients.
The statement being debated is that the best radiotherapy for treating prostate cancer should
involve hypofractionation. Taking together the current generally high rates of freedom from
biochemical failure achieved with modern conformal therapy using ~2 Gy fractions at (escalated)
total doses,19 the considerable theoretical doubts cast on the radiobiological modeling which
yielded  1.5 Gy, and the apparently good clinical outcomes recently reported from doing
exactly the opposite of hypofractionation, i.e., hyperfractionation, 16 I conclude that
48
hypofractionated radiotherapy for prostate cancer is contraindicated. The old adage "if it ain't
broke don't fix it" seems appropriate.
Rebuttal: John F. Fowler, D.Sc.
Dr. Nahum, you're not listening! Hypofractionation for prostate cancer came from biological
insight24 not a comparison of external-beam radiotherapy and low-dose-rate brachytherapy, and
this before there were other 5-y results. Since then, comparisons have all been at high dose-rate,
either with HDR brachytherapy (Brenner et al.,6 giving =1.2 Gy; Demanes et al.,17 1.2 Gy;
Martinez et al.,25 1.2 Gy) or with head-to-head linear accelerator trials (Lukka et al.,14 1.12 Gy;
Kupelian et al.,26 1.1 Gy).
Good clinical results are worth a hundred models. Modeling tumor TCP with hypoxia is
notoriously unreliable, with or without reoxygenation or inhomogeneities! The single clinical
exception mentioned by Dr. Nahum, has been dealt with by Bentzen and Ritter,15 who
demonstrated that the  value of 9.8 Gy from the Valdagni et al. nonrandomized clinical trial,16
had 95% confidence limits extending down to 0.7 Gy.14 Bentzen and Ritter suggested that
perhaps the two fractions/day control schedule of Valdagni et al. suffered from incomplete
repair. That was a penetrating comment because, if  were really low (instead of 8.3 Gy), it
would cause incomplete repair to contribute disproportionately more to the tumor effect, by
reason of the high repair capacity of prostate cancer cells. Then the control arm would have an
inflated effect, as reported.
Concerning in vitro values that showed a high  ratio, they are not relevant. Carlson et al.27
found that in vitro  values had little relationship to in situ values, and concluded that  for
prostate tumors is low.
If better tumor results can be obtained with the same or fewer complications, for half the number
of fractions or less17,25,26 than the conventional 1.8 or 2 Gy schedules, should we deny this unique
biological bonus to patients? This question is settled by the good clinical hypofractionated results
already quoted.
Rebuttal: Alan E. Nahum, Ph.D.
Professor Fowler is absolutely right when he asserts that hypofractionation is the "best" treatment
from the point of view of patients' convenience and healthcare economy (at least in the
macroeconomic sense; we are not discussing here the issue of individual clinics gaining more
income from prolonged, complex treatments such as multifractionated IMRT). But this is about
medical science not economics. It is not about patients' convenience. What we are debating is
whether hypofractionation is likely to lead to better clinical outcomes (i.e., uncomplicated cure
rates) than the presently practiced treatments at about 2 Gy/fraction. There are basically two
aspects to this issue, the theoretical one and the clinical results one.
Theoretical—any impartial observer would conclude that the theoretical case for a low prostate-
clonogen  and thus for hypofractionation is deeply flawed, relying on a paucity of clinical
data, in particular on one single published brachytherapy study—a case of comparing apples
(brachy) with oranges (external-beam).
49
Clinical—preliminary results from some new hypofractionation studies are emerging which do
not immediately kill stone dead the hypo-idea, but it is simply too early to draw definite
conclusions. Conversely there is an extensive and well-documented study of patients treated with
moderate hypofractionation (at about 3 Gy/fraction) from Manchester, UK (Ref. 9) which did not
demonstrate superior results and which is perfectly consistent with an  ratio of around 8 Gy.23
A study of hyperfractionation16 also yielded clinical outcomes inconsistent with a low .
But there are other considerations—if we currently fail to control a significant number of prostate
tumors due to hypoxia,22 then how much worse is this likely to be when much larger fractions are
used with the probable consequent impairment of reoxygenation between fractions?
In conclusion, hypofractionated radiotherapy undoubtedly has its place, especially when coupled
with excellent conformality, as the extracranial stereotactical treatment of early-stage lung
tumors amply demonstrates.28 Given the current generally excellent results using conformal
prostate radiotherapy at about 2 Gy/fraction, however, and the severe doubts cast on the
theoretical concept of a low , I conclude that to hypofractionate prostate treatments is to take
unnecessary risks with the health of patients suffering from prostate cancer.
REFERENCES
1. J. F. Fowler, M. A. Ritter, R. J. Chappell, and D. J. Brenner, "What hypofractionated

protocols should be tested for prostate cancer?" Int. J. Radiat. Oncol., Biol., Phys. 56,
1093–2003 (2003).
2. A. E. Nahum et al., "Incorporating clinical measurements of hypoxia into tumor local
control modeling of prostate cancer: implications for the  ratio," Int. J. Radiat. Oncol.,
Biol., Phys. 57, 391–401 (2003).
3. K. Haustermans and J. F. Fowler, "A comment on proliferation rates in human prostate
cancer," Int. J. Radiat. Oncol., Biol., Phys. 48, 29l (2000).
4. D. J. Brenner and E. J. Hall, "Fractionation and protraction for radiotherapy of prostate
carcinoma," Int. J. Radiat. Oncol., Biol., Phys. 43, 1095–1101 (1999).
5. J. F. Fowler, R. J. Chappell, and M. A. Ritter, "Is  for prostate tumors really low?" Int.
6. D. J. Brenner et al., "Direct evidence that prostate tumors show high sensitivity to
fractionation (low ratio) comparable to late-responding normal tissue," Int. J. Radiat.
Oncol., Biol., Phys. 52, 6–13 (2002).
7. J. Z. Wang, M. Guerrero, and X. A. Li, "How low is the  ratio for prostate cancer?"
8. C. D. Collins, R. W. Lloyd-Davies, and A. V Swan, "Radical external beam radiotherapy
for localized carcinoma of the prostate using a hypofractionation technique," Clin. Oncol.
3, 127–132 (1991).
9. J. E. Livsey et al., "Hypofractionated conformal radiotherapy in carcinoma of the
prostate: five-year outcome analysis," Int. J. Radiat. Oncol., Biol., Phys. 57, 1254–1259
(2003).
10. T. Djemil, C. A. Reddy, T. R. Willoughby, and P. Kupelian, "Hypofractionated intensity-
modulated radiotherapy (70 Gy at 2.5 Gy per fraction) for localized prostate cancer,"
Int. J. Radiat. Oncol., Biol., Phys., Suppl. 57, S275–S276 (2003).
11. Y. Yoshioka et al., "High-dose-rate brachytherapy as monotherapy for localized prostate
cancer: A retrospective analysis with special focus on tolerance and chronic toxicity," Int.
50
12. B. L. Madsen et al., "Intra-fractional stability of the prostate using a stereotactic

radiotherapy technique," Int. J. Radiat. Oncol., Biol., Phys. 57, 1285–1291 (2003).
13. G. Gustafson et al., "High dose rate (HDR) monotherapy for early stage prostate cancer:
toxicity results utilizing the common toxicity criteria," Int. J. Radiat. Oncol., Biol., Phys.
57, S230–S231 (2003).
14. H. Lukka et al., "A randomized trial comparing two fractionation schedules for patients
with localized prostate cancer," Radiother. Oncol. 69, (Supplement 1), S7 (2003)
(ScienceDirect).
15. S. M. Bentzen and M. A. Ritter, "The  ratio for prostate cancer: what is it, really?"
Radiother. Oncol. 76, 1–3 (2005).
16. R. Valdagni et al., "Standard or hyperfractionated conformal radiation therapy in prostate
cancer: A prospective non-randomized trial comparing standard and hyperfractionated
conformal radiation therapy," Radiother. Oncol. 75, 74–82 (2005).
17. D. J. Demanes et al., "High-dose-rate intensity-modulated brachytherapy with external
beam radiotherapy for prostate cancer: California Endocurietherapy's 10-year results,"
18. J. F. Fowler et al., "Analysis of 5y clinical results of EBRT + brachytherapy shows that
 is low," Int. J. Radiat. Oncol., Biol., Phys. (submitted).
19. G. E. Hanks et al., "Dose response in prostate cancer with 8–12 years' follow up," Int. J.
20. R. G. Stock et al., "A dose-response study for I-125 prostate implants," Int. J. Radiat.
Oncol., Biol., Phys. 41, 101–108 (1998).
21. C. R. King and J. F. Fowler, "A simple analytic derivation suggests that prostate cancer
 ratio is low," Int. J. Radiat. Oncol., Biol., Phys. 51, 213–214 (2001).
22. B. Movsas et al., "A hypoxic ratio of prostate pO2/muscle pO2 predicts for biochemical
failure in prostate cancer patients." Urology 60, 634–639 (2002).
23. A. E. Nahum and J. D. Chapman, "Dose-escalation strategies for increasing prostate local
control: how clinically proven hypoxia radically changes the picture," Radiother. Oncol.
68, S113 (2003) (ScienceDirect).
24. G. M. Duchesne and L. J. Peters, "What is the ratio for prostate cancer? Rationale for
hypofractionated high-dose-rate brachytherapy," Int. J. Radiat. Oncol., Biol., Phys. 44,
747–748 (1999).
25. A. A. Martinez et al., "Lack of benefit from a short course of androgen deprivation for
unfavorable prostate cancer patients treated with an accelerated hypofractionation
regime," Int. J. Radiat. Oncol., Biol., Phys. 62, 1322–1331 (2005).
26. P. A. Kupelian et al., "Hypofractionated intensity-modulated radiotherapy (70 Gy at
2.5 Gy per fraction) for localized prostate cancer," Int. J. Radiat. Oncol., Biol., Phys.
63(5), 1463–1468 (2005).
27. D. J. Carlson et al., "Comparison of in vitro and in vivo  ratios for prostate cancer,"
Phys. Med. Biol. 49, 4477–4491 (2004).
28. J. F. Fowler et al., "A challenge to traditional radiation oncology," Int. J. Radiat. Oncol.,
Biol., Phys. 60(4), 1241–1256 (2004).
51
1.11. The routine use of personal patient dosimeters is of little value

in detecting therapeutic misadministrations
Arnold Feldman and Fredric Marc Edwards
OVERVIEW
Analytical models have traditionally been used to estimate dose distributions for treatment
planning in radiation therapy. Recently, some physicists have suggested that Monte Carlo
techniques yield more accurate computations of dose distributions, and a few vendors of
treatment planning systems have incorporated Monte Carlo methods into their software. Other
physicists argue that, for a number of reasons, analytical methods should be preserved. This
controversy is the topic of this Point/Counterpoint article. Thanks are extended to Paul Nizin,
Ph.D. of Baylor College of Medicine for suggesting the topic.
Arguing for the Proposition is Radhe Mohan, Ph.D. Dr. Mohan received his Ph.D. from Duke
University and is currently Professor and Director of Radiation Physics at the Medical College of
Virginia (MCV) Hospitals, Virginia Commonwealth University. Dr. Mohan has been actively
engaged in research and clinical implementation of advanced dose calculation methods, 3D
treatment planning, Monte Carlo techniques and IMRT for 25 years, first at Memorial Sloan-
Kettering Cancer Center and now at MCV. He has published and lectured widely on these topics
at national and international meetings and symposia.
Arguing against the proposition is John Antolak, Ph.D. Dr. Antolak received his Ph.D. in
Medical Physics from the University of Alberta (Canada) in 1992. He then joined the Department
of Radiation Physics at The University of Texas M. D. Anderson Cancer, where he is currently
an Assistant Professor. He is certified by the American Board of Radiology and licensed to
practice Medical Physics in Texas. He is active in the education of graduate students,
dosimetrists, and other physicists, and his research interests center around the use of electron
beams for conformal radiotherapy. In his spare time, he enjoys playing ice hockey and coaching
his son's ice hockey team.
FOR THE PROPOSITION: Radhe Mohan, Ph.D.
Opening Statement
Monte Carlo techniques produce more accurate estimates of dose than other computational
methods currently used for planning radiation treatments. Were it not for limitations of computer
speed, Monte Carlo methods probably would have been used all along.
With the spectacular increase in computer speed and the development of clever algorithms and
variance reduction schemes, Monte Carlo methods are now practical for clinical use. The time
required to compute a typical treatment plan has shrunk to a few minutes on computers costing
52
less than $50,000. A few centers have started using Monte Carlo techniques for clinical purposes,
and releases of commercial products are imminent.
As with any new product, an "adjustment period" will be needed during which we learn how to
apply this powerful tool. Some find the "statistical jitter" in Monte Carlo results troubling. This
issue is being addressed by several investigators. The additional cost of hardware and software
may be another obstacle, but is likely to be resolved as computers become cheaper and more
powerful.
Another issue is whether improvements in accuracy are clinically significant and worth the
additional cost. It is difficult to answer the first question unequivocally, because randomized
trials in which half the patients are treated with less accurate methods are not feasible. Arguments
in favor of using Monte Carlo methods include:
(1) Elimination of the need to continually reinvent approximate dose computation models and to
tweak them to meet every new situation, as well as the need for trial and error approaches to
obtain acceptable matches with measured data. The medical physics community has been
engaged in such exercises for 50 years. It is time to stop.
(2) Broad applicability and accuracy of the same Monte Carlo model for all anatomic geometries
and treatment modalities (photons, electrons, brachytherapy). With analytical methods, there is a
separate model for each modality and a unique set of approximations and assumptions is required
for each type of field shaping device.
(3) Dramatic reduction in the time, effort and data required for commissioning and validating the
dose computation part of treatment planning systems.
(4) Improved consistency of inter-institutional results, and greater quality of dose response data
because of improved dose accuracy.
(5) Accurate estimation of quantities difficult or impossible to measure, such as dose distributions
in regions of disequilibrium.
Until recently, the major reason for considering Monte Carlo methods was the inaccuracy of
semi-empirical models for internal inhomogeneities and surface irregularities. Now an equally
important justification is the ability of Monte Carlo techniques to provide accurate corrections for
transmission through, scattering from, and beam hardening by field shaping devices. Monte
Carlo techniques are also able to account correctly for radiation scattered upstream from field-
shaping devices. These effects are quite significant for small fields encountered in intensity-
modulated radiotherapy.
The transition to Monte Carlo methods will have to be gradual. Even though a few minutes of
time to compute a plan may seem insignificant, computer-aided optimization of treatment plans
may require many iterations of dose computations. In these situations, hybrid techniques will be
needed that use fast but less accurate conventional models for most optimization iterations and
Monte Carlo techniques for the remainder.
Since Monte Carlo techniques are now affordable and practical, there is no reason not to use
them. It is not necessary to conduct clinical trials to once again prove the clinical significance of
improved dose accuracy. Monte Carlo methods should be deployed in radiation therapy with
53
deliberate speed. For some applications, such as, IMRT optimization, it may be necessary to
continue to use high-speed conventional methods in conjunction with Monte Carlo techniques at
least for now.
Rebuttal
Dr. Antolak has raised several issues, some of which were addressed in my Opening Statement.
With faster computers and clever schemes to reduce variance, the stochastic nature of the Monte
Carlo approach is no longer an impediment. Statistical uncertainty of 1%–2% is achievable on
grid sizes of 2–3 mm in MC dose distribution calculations, requiring just a few minutes on easily
affordable multiprocessor systems. While statistical noise may be unsightly, its effect on the
evaluation of dose-volume and dose-response parameters of plans is insignificant. In addition,
techniques to smooth out noise are being implemented.
Analytic models introduce systematic errors in dose. They simply cannot achieve the accuracy of
MC techniques. While it is true that analytic models consistently produce precise results for the
same input data, these results are consistently inaccurate.
Dr. Antolak is concerned that approximations to speed up Monte Carlo computations may affect
the accuracy of results. But Monte Carlo developers and users should always ensure that
approximations have no significant impact on accuracy. Nothing else should be necessary.
Responses to other such concerns raised by Dr. Antolak are: (1) Considering the uncertainties in
dose-response information and other sources of data in the radiotherapy chain, our ability to
define "how much noise in the dose distributions is acceptable" is similar to our ability (or lack
thereof) to determine the level of dose inaccuracy that may be acceptable. (2) Dose to a point is
not a meaningful quantity when Monte Carlo techniques are used. Beam weighting and dose
prescription should be specified in terms of dose to fractional volumes (e.g., 98% of the tumor
volume). (3) Statistical noise should have practically no effect on inverse treatment planning
because the intensity along a ray is affected by the average of dose values over a large number of
voxels lying along the ray and not by the dose in any one voxel. (4) Commissioning of Monte
Carlo algorithms will be the responsibility of the same physicists and/or commercial vendors who
commission conventional methods.
I believe strongly that concerns raised by Dr. Antolak and others are being resolved and that we
are now ready to introduce Monte Carlo techniques into clinical use.
AGAINST THE PROPOSITION: John Antolak, Ph.D.
Opening Statement
We have a professional responsibility to ensure that patient treatments are accurately delivered,
and the accuracy of treatment planning dose computation is one aspect of this. There are data to
support the conclusion "that Monte Carlo techniques yield more accurate computations of dose
distributions," provided that the Monte Carlo technique is fully applied. However, in light of
other factors detailed below, Monte Carlo methods should not replace analytical methods for
estimating dose distributions.
54
Before arguing against the proposition, it is necessary for me to clarify what I believe the basic
difference is between an analytical method and a Monte Carlo method. It boils down to the
difference between deterministic and stochastic. The Monte Carlo method is stochastic, i.e.,
independent calculations of the same problem will give different answers. The analytical method
is deterministic, i.e., independent calculations of the same problem will give the same answer, at
least to within the limits of numerical round-off and truncation errors. In my opinion and for the
purpose of this discussion, any nonstochastic method is considered to be an analytical method.
The accuracy of an algorithm (or method) describes how close it comes to the true answer.
Clinical physicists have to worry about the accuracy of both analytical and Monte Carlo methods.
The full Monte Carlo method (e.g., EGS4) is considered by many to be the gold standard for
accurate dose calculations. The precision of an algorithm is a measure of the repeatability of the
answer. Analytical methods have essentially absolute precision. However, the precision of the
Monte Carlo method, as measured by the standard error, is proportional to the inverse of the
volume of the dose voxels, and to the inverse square root of the computational resources
allocated to the problem. For example, reducing the standard error by a factor of two requires
four times as much CPU-time. Variance reduction techniques can be used to reduce the
computational resources required to obtain a given precision. However, the time (or resources)
required for full Monte Carlo simulations of patient dose distributions is currently too great for
clinical use. By necessity, current Monte Carlo treatment planning algorithms (those being touted
for clinical use) introduce approximations that greatly speed up the calculations, but the accuracy
of the results may be affected. At the same time, significant improvements are also being made to
the accuracy of analytical algorithms. Also, for the clinical physicist, commissioning analytical
algorithms is relatively straightforward, noise is not a problem, and the accuracy can be easily
documented.
From the perspective of the clinical physicist, many questions about the use of Monte Carlo
algorithms have not yet been answered. How much noise in the dose distributions is acceptable?
In the presence of noise, how should beam weighting (e.g., isocentric weighting) be done? What
effect does noise have on inverse treatment planning? Who will take responsibility for
commissioning the algorithm, and how accurate are the results of the commissioning? How long
will the calculation take relative to faster analytical calculations? How will the calculation time
affect treatment-planning throughput, particularly when using optimization methods? Is the
spatial resolution sufficient for clinical use? Most of the time, Monte Carlo treatment planning
calculation times are quoted for relatively coarse (e.g., 5 mm) spatial resolution. Just reducing the
resolution from 5 mm to 3 mm requires approximately five times as much CPU-time. These are
just some of the issues that need to be resolved and well-documented before Monte Carlo
methods can replace analytical methods for treatment planning.
Monte Carlo methods may be used as an independent verification of the dose delivery, or to
document (rather than plan) the dose delivery. However, until the questions above are
successfully answered, Monte Carlo methods should not replace analytical methods for
estimating radiotherapy dose distributions.
Rebuttal
We agree that greater accuracy in dose computation is desirable, Monte Carlo methods can
produce more accurate dose estimates, and "Monte Carlo methods should be deployed in
radiation therapy with deliberate speed." However, these points are not the proposition we are
addressing.
55
A potential patient recently inquired about the status of Monte Carlo planning at our institution.
From what he had read, he believed that Monte Carlo treatment planning is a "silver bullet." Dr.
Mohan says it is time to stop reinventing. I believe that implementation of "clever algorithms and
variance reduction schemes" is reinventing Monte Carlo treatment planning methods. Further,
trial and error will not stop with Monte Carlo. With complete information about source and
machine geometry, Monte Carlo calculations can be highly accurate. However, Monte Carlo
algorithms usually start with a source model that requires trial and error adjustments to match
measured data.
Whereas reinventing analytical methods usually improves accuracy, reinventing Monte Carlo
methods may decrease accuracy. In both cases, there is a tradeoff between accuracy and speed,
which is often seen if the Monte Carlo approach averages the dose over large voxels. How is the
accuracy of a particular implementation judged and to what should it be compared? The
"spectacular increase in computer speed and the development of clever algorithms" noted by Dr.
Mohan permits significant improvements in analytical models, potentially leading to a model for
coupled photon-electron transport under more general conditions.
Future reductions in commissioning and validation efforts will only come from manufacturers'
standardization of treatment machines and improved quality in their construction. Modified and
new machines will still require extensive commissioning and validation for both Monte Carlo and
analytical methods. Considerable research remains to be done to identify the minimum data set
sufficient to validate input data that characterizes a treatment machine. Dr. Mohan's last two
points are really arguments for greater dose accuracy and apply to treatment planning systems in
general, not just Monte Carlo methods.
Dr. Mohan cites the significance of Monte Carlo methods applied to field shaping and intensity
modulation devices. These applications can be very complex, but are usually not modeled
explicitly. For example, modeling all of the field segments for a complex DMLC fluence pattern
is impractical under normal circumstances. Using an approximate approach affects the overall
accuracy of the dose calculation, as it would for an analytical method.
Monte Carlo methods are an invaluable tool for improving analytical models to a point where
their dose uncertainty is insignificant compared with other uncertainties radiation therapy—such
as setup, internal organ motion, target delineation, and biological response. As stated earlier,
"Monte Carlo methods may be used as an independent verification of dose delivery, or to
document (rather than plan) dose delivery," but should not replace analytical methods for
estimating dose distributions in radiotherapy treatment planning.
56
1.12. Thermoradiotherapy is underutilized for the treatment of

cancer
Eduardo G. Moros and Peter M. Corry
Reproduced from Medical Physics, Vol. 34, No. 1, pp. 1–4, January 2007
OVERVIEW
Combining heat (hyperthermia) with radiotherapy (thermoradiotherapy) has a long and

somewhat checkered past. There is no question that heat enhances the effect of radiation, but it
does so for both cancerous and normal tissues. A number of clinical trials were conducted by the
Radiation Therapy Oncology Group in the 1980s and 1990s in the hope of demonstrating that
hyperthermia could improve the therapeutic ratio, but the results were not promising enough to
encourage widespread interest in thermoradiotherapy by either radiation oncologists or funding
agencies. Consequently, at least in the USA, support for hyperthermia research dwindled and
reimbursement for hyperthermia treatments was curtailed. Recently, however, promising results
have been published on several clinical trials in Europe. This has rekindled interest in
thermoradiotherapy and some have suggested that valuable opportunities for the use of
hyperthermia are being missed and that hyperthermia is being underutilized. Others disagree.
This is the premise debated in this month's Point/Counterpoint debate.
Arguing for the Proposition is Eduardo G. Moros, Ph.D. Dr. Moros received his B.S., M.S. and
Ph.D. degrees in Mechanical Engineering from the University of Arizona, Tucson, in 1984, 1987,
and 1990, respectively. His graduate research in bioheat transfer and acoustic propagation
modeling for hyperthermia therapy applications brought him into the field of Radiation Oncology
in 1985. After receiving his Ph.D., Dr. Moros spent a year as an Associate Researcher at the
University of Wisconsin at Madison and then joined the Mallinckrodt Institute of Radiology at
Washington University School of Medicine, where he eventually became a Full Tenure Professor
and Head of the Research Physics Section of the Department of Radiation Oncology. In 2005, Dr.
Moros accepted a position at the University of Arkansas for Medical Sciences, Little Rock, AR,
as the Director of the Division of Radiation Physics and Informatics of the Department of
Radiation Oncology. His major research activities are in noninvasive temperature estimation
using ultrasound, radiosensitivity by the cellular stress response, imaging of hyperthermia-
induced tumor oxygenation using positron emission tomography and thermal ablation, and the
development of clinical devices for thermoradiotherapy, hyperthermia, bioelectromagnetics, and
bioheat transfer. He is a Past-President of the Society for Thermal Medicine and a member of the
Board of Editors of Medical Physics and the International Journal of Hyperthermia.
Arguing against the Proposition is Peter M. Corry, Ph.D. Dr. Corry is presently Distinguished
Professor and Vice Chairman in the Department of Radiation Oncology at the University of
Arkansas for Medical Sciences in Little Rock, AR. From 1986 to 2006 he held the positions of
Director of Medical Physics and Director, Radiation Oncology Research Laboratories at the
William Beaumont Hospital in Royal Oak, MI. Prior to that (1968–1986) he was Professor of
Biophysics in the Department of Physics at the University of Texas M. D. Anderson Cancer
Center in Houston Texas. He was certified by the American Board of Radiology in Radiological
57
Physics in 1965. Dr. Corry was recently elected President-Elect of the Radiation Research
Society.
FOR THE PROPOSITION: Eduardo G. Moros, Ph.D.
Opening statement
The following statements are restricted to thermal medicine with mild temperatures (39–43 °C),
commonly known as hyperthermia. Supported by a wealth of biological data and encouraging
initial clinical trials, hyperthermia was embraced by many in the radiation oncology community
in 1970s and 1980s as a potent adjunct anticancer modality. 1 New companies produced clinical
systems and those for superficial and interstitial therapy quickly received Food and Drug
Administration (FDA) clearance. By 1984, the first hyperthermia CPT codes were approved,
which indicates a level of general acceptance by practitioners. The enthusiasm was such that the
Radiation Therapy Oncology Group (RTOG) organized several clinical trials intended to
scientifically establish the effectiveness of thermoradiotherapy. To the surprise of most and the
demise of a promising new modality, the results of these clinical trials were negative. The main
deficiency was quickly identified as ―Quality Assurance‖ (QA) as evidenced in the titles of
several papers reporting on the RTOG trials.2 I evaluated the QA data for the RTOG-8419 phase
III clinical trial that compared interstitial thermoradiotherapy with interstitial radiotherapy of
recurrent or persistent tumors.3 To give readers an appreciation of the problem, my two main
observations were: (1) the treatment volume defined by the thermobrachytherapy implants was
almost always smaller than the reported tumor volume, and (2) the recorded temperatures in the
thermoradiotherapy arm (86 patients) revealed that only one patient had an adequate
hyperthermia session. The first finding is indicative of poor volumetric tumor coverage for both
modalities while the second argues that adequate hyperthermia was never given. Results from
both arms were identical!
The RTOG addressed the QA deficiencies with QA guidelines that were published in a series of
papers in the International Journal of Radiation Oncology, Biology, Physics and a new protocol
for recurrent breast cancer was initiated. However, by that time enthusiasm had waned and all
eyes were being fixed understandably on three-dimensional conformal radiotherapy and other
new technologies. In North America, only a small number of clinics continued to practice
hyperthermia, and even a smaller number of academic groups continued hyperthermia research.
Although the negative news affected Europeans likewise, they managed to perform several phase
III clinical trials with emphasis on adhering to internationally accepted QA guidelines. All these
European trials have been positive.4 The importance of QA cannot be overemphasized as it has
been the main factor differentiating negative from positive trials.5,6
The lesson learned is that hyperthermia is a potent adjunct modality to ionizing radiation,
doubling response rates in most cases and extending survival in some cases, if administered as
prescribed.7 Furthermore, the main synergistic effects of hyperthermia when combined with
radiotherapy and/or chemotherapy—such as inhibition of radiation damage repair,
radiosensitization, perfusion changes, reoxygenation, the universal heat shock response, and
immunologic effects—require relatively mild and clinically achievable temperatures around
41.5 °C.8 In conclusion, the growing body of scientific data in support of hyperthermia, the
positive phase III clinical trials when adequate QA guidelines are followed, and the remarkable
58
benefits that it offers to patients, clearly and strongly support the proposition that
thermoradiotherapy is currently underutilized in the treatment of cancer.
AGAINST THE PROPOSITION: Peter M. Corry, Ph.D.
Opening statement
Thermoradiotherapy is not presently underused in clinical application for a variety of reasons. In

the first instance lower temperature hyperthermia (40–43 °C) has no long-term efficacy when
used as single agent therapy. These observations have led to the conclusion that hyperthermia
must be applied in combination with other forms of anticancer therapy such as ionizing radiation
or chemotherapy. This factor alone decreases the utility of hyperthermia in the armamentarium of
cancer treatments.
Over the past decade or so there have been a number of reports in the literature that hyperthermia
has some benefit when combined with conventional radiation therapy (thermoradiotherapy) 5,9,10
but these studies have mostly addressed the issue of superficial recurrent, metastatic or advanced
tumors. There are no reports of its application to the treatment of potentially curable cancers
where a significant impact might be possible. Even in the cases where some benefit was claimed,
it was often marginal. For example, in the study by Sneed et al.11 treating recurrent Glioblastoma
Multiforme (GBM) with interstitial radiation plus interstitial hyperthermia, it was found that the
median survival was only modestly increased from 76 to 85 weeks by the addition of heat to the
radiation regimen. Is it worth all the extra effort and expense to add heat to the radiation
regimen? Another recent study combining heat with radiation and chemotherapy
(thermochemoradiotherapy) for cancer of the uterine cervix was carried out in several institutions
in Europe and North America.12 Unfortunately the reproducibility of the results was less than
optimal. In Europe the disease free survival rate was approximately 80% while in North America
it bottomed out at approximately 40%. Certainly better reproducibility is required before
widespread clinical application is attempted.
Another impediment to widespread clinical application is the paucity of available equipment

types and manufacturers for the hyperthermia aspect of thermoradiotherapy. At the present time
there are only two manufacturers that sell equipment for FDA approved clinical application in the
United States, one using microwaves and the other employing planar ultrasound technology.
Neither of these systems is of any utility for other than small superficial tumors (about 3–4 cm)
most of which could be easily excised surgically. There are no FDA approved systems available
for treating deep seated tumors although some experimental systems are undergoing evaluation.
However, none of these systems is able to noninvasively treat tumors in the thorax where the
treatment of lung cancer is desperately in need of improvement. There are no systems even on the
horizon for this latter application. A final limitation is reimbursement for services rendered
which, at the present time, is insufficient to cover the cost of equipment acquisition let alone the
myriad of other expenses involved. In general the state of affairs with thermoradiotherapy seems
to favor its use in large academic centers where research and development can continue both into
the evolution of new systems that can more effectively apply it, as well as investigation into its
clinical efficacy.
Rebuttal: Eduardo G. Moros, Ph.D.
Following are my responses to the five major issues raised by Dr. Corry.
59
Hyperthermia is not an effective single therapy. Early studies showed that hyperthermia alone
does not have the lasting effectiveness of mainline therapies (i.e., surgery, chemotherapy and
radiotherapy). However, in the fight against cancer combined therapies are the norm.
Furthermore, many new molecular therapeutics seek to enhance mainline therapies rather than
replace them.13,14 If a new adjuvant therapy were to produce a doubling of the local control rate
and a significant extension of time to recurrence, it would be praised as a breakthrough.
Ironically, this is exactly the effect of thermoradiotherapy for several cancers. 1,4,5,7
No data on curable cancers. As with most new therapies, hyperthermia has been tested clinically
with aggressive, recurrent, and persistent cancers that fail to be cured by standard treatments.
Thus, Dr. Corry's observation that there is an absence of reports on potentially curable cancers is
not surprising. Nevertheless, given that the addition of hyperthermia has been shown to have a
significant impact on the treatment of tumors known to be resistant to other therapies, one can
reasonably conclude that administering hyperthermia early in the management of cancer patients
is warranted. The fact that the addition of only 60 min of hyperthermia, 30 min before and after
a brachytherapy implant boost, prolonged the survival of patients with GBM should be a strong
incentive for expanding the use of hyperthermia to curable cancers.11 Is there any other agent that
has produced such a positive result with GBMs?
Reproducibility. This is an issue for all therapies and in particular for radiotherapy and
hyperthermia. Reproducibility involves complex issues of clinical trial design, patient selection,
treatment technique, dosage, sequencing of therapies, quality assurance, etc., that need the expert
oversight of cooperative groups like the RTOG. To avoid missing the forest from the trees,
however, a look at a summary of clinical trials4 clearly shows that the net hyperthermic benefit is
highly reproducible.
Paucity of equipment and reimbursement. I agree completely with Dr. Corry on this issue. There
are three main intertwined factors affecting innovation in hyperthermia. First is the low clinical
interest after the negative trials of the 1980s. Second is that reimbursement levels have remained
frozen for more than a decade due to the lack of significant clinical volume nationwide. In recent
times, payments for hyperthermia have increased, but the new amounts are still too low. Third
factor is the 1976 Medical Device Amendments law, which does not allow hyperthermia
manufacturers the use of the 510(k) premarket notification for FDA clearance. (Surprisingly,
radio frequency ablation devices, which use much higher temperatures to destroy tissues, can use
the predicate device route to obtain FDA approval.) Consequently, manufacturers have to invest
considerable resources and years of clinical trials to show safety and efficacy using
Investigational Device Exemptions. This situation discourages private investment in new
hyperthermia technologies that could only improve the hyperthermic effect. One last thing—it is
often said that hyperthermia is too time consuming and manpower intensive. I posit that this
argument will disappear completely the moment hyperthermia treatments are adequately
reimbursed.
Rebuttal: Peter M. Corry, Ph.D.
In his Opening Statement, Dr. Moros presented an admirable account of the difficulties
encountered in applying hyperthermia in routine clinical application. He did a comprehensive job
of discussing the deficiencies associated with the RTOG clinical trials 2,3 that drastically reduced
enthusiasm for, and all but eliminated, clinical hyperthermia applications in North America and,
to a lesser extent, in Europe and Asia. He devoted approximately 2/3 of his opening statement to
QA issues of early trials. The RTOG attempted to address these shortcomings by convening
60
study groups that published general QA guidelines for clinical hyperthermia (e.g., see Ref. 15) as
well as three specific guidelines for ultrasound, deep seated tumors and interstitial hyperthermia.
These guidelines were published in the early 1990s in the International Journal of Radiation
Oncology, Biology, Physics. Where Dr. Moros went wrong in his argument was to imply that this
effort solved all the QA issues associated with the application of clinical hyperthermia, and that
we were now on the road to blissful harmony with our thermal measuring and administration
devices.
The published guidelines from the RTOG require extensive thermometry, usually in three
dimensions, as just one of the requirements for clinical application. There are also several other
requirements for applicator selection, particularly relating to adequate tumor coverage. The rub
here is that none of the clinical trials cited by Dr. Moros or me complied strictly with those rather
rigid guidelines. Some clinical programs continue to apply ―hyperthermia‖ without measuring
intratumoral temperature. Other programs obtain one or two temperature points on day one and
assume that, with the given power settings for the equipment, the temperatures will be
reproducible throughout the course of treatment. Both of these assumptions have been shown to
be clearly wrong.16 Just a few sensors (1–3) always give a more pleasing picture (falsely) than a
larger number, say 10–12. It was also shown that, as one might expect, things change during the
course of treatment and that serial temperature measurements are much more enlightening. In
short, there is so much variation in the QA aspects of clinical trials published that it is often
difficult to evaluate results and sometimes impossible to compare them even among leading
institutions in the field.
Why do these problems persist after decades of research and development? There is a simple
one-word answer: money. In order to comply with the published thermometry guidelines
sophisticated MR thermometry systems are required. This raises the cost of a hyperthermia
system by $2–3 million—not to mention facility construction costs, the fact that MR compatible
heating systems, which are not widely available, are required, and that a cadre of capable
scientists is needed to make sure that everything is running as it should. I discussed in some detail
the financial constraints imposed by governments, particularly in North America, in my Opening
Statement. They are formidable and challenging. Exacerbating the situation is the drought that we
are currently experiencing in the North American research funding arena, which is hobbling
more advanced programs in academia. All of this adds up to a rather grim outlook for
hyperthermia applications in the near future.
REFERENCES
1. R. J. Myerson, E. G. Moros, J. L. Roti Roti, ―Hyperthermia,‖ in Principles and Practice

of Radiation Oncology, edited by C. A. Perez and L. W. Brady (Lippincott-Raven, Philadelphia,
1998), pp. 637–683.
2. C. A. Perez et al., ―Quality assurance problems in clinical hyperthermia and their impact
on therapeutic outcome: A report by the Radiation Therapy Oncology Group,‖ Int. J. Radiat.
Oncol., Biol., Phys. 16, 551–558 (1989).
3. B. Emami et al., ―Phase III study of interstitial thermoradiotherapy compared with
interstitial radiotherapy alone in the treatment of recurrent or persistent human tumors. A
prospectively controlled randomized study by the Radiation Therapy Oncology Group,‖ Int. J.
4. M. W. Dewhirst and P. K. Sneed, ―Those in gene therapy should pay closer attention to
lessons from hyperthermia,‖ Int. J. Radiat. Oncol., Biol., Phys. 57, 597–599.
61
5. E. L. Jones et al., ―Randomized trial of hyperthermia and radiation for superficial

tumors,‖ J. Clin. Oncol. 23, 3079–3085 (2005).
6. E. L. Jones et al., ―In regard to Vasanathan et al. (Int. J. Radiat. Oncol., Biol., Phys. (61,
145–153, 2005),‖ Int. J. Radiat. Oncol., Biol., Phys. 63, 644 (2005).
7. M. H. Falk and R. D. Issels, ―Hyperthermia in oncology,‖ Int. J. Hyperthermia 17, 1–18
(2001).
8. M. W. Dewhirst et al., ―Re-setting the biologic rationale for thermal therapy,‖ Int. J.
Hyperthermia 21, 779–790 (2005).
9. J. Overgaard et al., ―Randomized trial of hyperthermia as adjuvant to radiotherapy for
recurrent or metastatic malignant melanoma: European Society for Hyperthermic Oncology,‖
Lancet 345, 540–543 (1995).
10. C. C. Vernon et al., ―Radiotherapy with or without hyperthermia in the treatment of
superficial localized breast cancer: Results from five randomized controlled trials—International
Collaborative Hyperthermia Group,‖ Int. J. Radiat. Oncol., Biol., Phys. 35, 731–744 (1996).
11. P. K. Sneed et al., ―Survival benefit of hyperthermia in a prospective randomized trial of
brachytherapy boost + hyperthermia for Glioblastoma Multiforme,‖ Int. J. Radiat. Oncol., Biol.,
Phys. 40, 287–295 (1998).
12. A. M. Westermann et al., ―First results of triple-modality treatment combining
radiotherapy, chemotherapy, and hyperthermia for the treatment of patients with stage IIB, III
and IVA cervical carcinoma,‖ Cancer 104, 763–769 (2005).
13. D. Cerna, K. Camphausen, and P. J. Tofilon, ―Histone deacetylation as a target for
radiosensitization,‖ Curr. Top. Dev. Biol. 73, 173–204 (2006).
14. D. W. Kim et al., ―Molecular strategies targeting the host component of cancer to
enhance tumor response to radiation therapy,‖ Int. J. Radiat. Oncol., Biol., Phys. 64(1), 38–46
(2006). M. W. Dewhirst et al., ―RTOG quality assurance guidelines for clinical trials using
hyperthermia,‖ Int. J. Radiat. Oncol., Biol., Phys. 18, 1249–1259 (1990).
15. P. M. Corry et al., ―Evaluation of equipment for hyperthermic treatment of cancer,‖ Int. J.
Hyperthermia 4, 53–74 (1988).
62
CHAPTER 2
Highly Conformal Radiotherapy: IMRT, Tomotherapy,

Stereotactic Radiosurgery, Proton Therapy
2.1. It is necessary to validate each individual IMRT treatment plan
before delivery
Chester Ramsey and Scott Dube

Reproduced from Medical Physics, Vol. 30, No. 9, pp. 2271–2273, September 2003
OVERVIEW
Many physicists take the position that IMRT treatment plans are complex and must be validated
before use because small errors can adversely affect patient treatment. These physicists feel that
the time devoted to validation is completely justifiable. Other physicists believe that such
validation can be eliminated, or at least substantially streamlined, if appropriate dosimetric and
quality assurance measures are deployed by the physicist. They argue that validation of
individual IMRT treatment plans is a misuse of time and resources. This difference in perspective
is addressed in this month's Point/Counterpoint.
Arguing for the Proposition is Chester R. Ramsey, Ph.D. Dr. Ramsey is Director of Medical
Physics at the Thompson Cancer Survival Center in Knoxville, Tennessee. He received an M.S.
(1996) and Ph.D. (2000) from the University of Tennessee and is board certified in radiation
oncology physics (ABMP). He is author or co-author on 12 journal articles and over 75
presentations and has been the instructor of multiple IMRT training courses. Dr. Ramsey is a
member of the Medical Physics Editorial Board, AAPM Task Group 76, AAPM Economics
Committee, and ACR Medical Physics Economics Committee.
Arguing against the Proposition is Scott Dube, M.S. Scott Dube graduated from the University of
Colorado Medical Center with an MS in Radiological Sciences in 1979. His first job was
working for Rocky Mountain Medical Physics in Denver. Soon he was off to Hawaii to join Mid-
Pacific Medical Physics. In 1987 he moved to Oregon and worked for the Northwest Medical
Physics Center but returned a year later to be Senior Medical Physicist at The Queen's Medical
Center in Honolulu. Although Scott resides in the middle of the Pacific Ocean, he feels very
connected to the worldwide radiation therapy community through his participation in various
listservers, such as medphys, pinnacle, and impac.
FOR THE PROPOSITION: Chester Ramsey, Ph.D.
Opening Statement
Over the past two years, the number of patients treated with IMRT has been growing at an
exponential rate. This growth is due in large part to the approval of IMRT reimbursement by
63
Medicare, and to the associated "IMRT-hype" from radiotherapy vendors. IMRT has been shown
to have the potential to improve outcomes for multiple disease sites and pathologies.1,2 However,
the clinical success or failure of each institution's IMRT program depends on the correct delivery
of 3D dose distributions calculated by the planning system (Planned Dose) to the correct location
in the patient (Delivered Dose).
There are many potential errors that can arise during IMRT planning and delivery. It is the
responsibility of the Medical Physicist to ensure that the Planned Dose "agrees" with the
Delivered Dose, and to address any issues arising from geometric or dosimetric errors for each
IMRT patient. Unlike conventional and 3D conformal radiotherapy, IMRT plans must be verified
for each individual patient because of the many sources of error.
The difference between Planned Dose and Delivered Dose (i.e., Error) in IMRT comes from
three sources: (1) Treatment Planning, (2) Treatment Delivery Mechanics, and (3) Time-
Dependent Target/Tissue Positioning. Even if all errors are eliminated in the treatment planning
process using a complete Monte Carlo based model, it would still be impossible to account for
downstream errors without individual verification.
IMRT delivery systems have inherent mechanical errors due to leaf transit time (tomotherapy-
based systems) or leaf positional error (MLC based systems). 3,4,5 Likewise, the accuracy of
compensator-based IMRT is limited by the accuracy of the milling equipment. In multiple static
segmental and dynamic IMRT, dose errors can occur from the rapid "on and off" of the x-ray
field.6 Calibrations can drift on motor encoders, leaf offsets can change during MLC
maintenance, and gravitational sagging can effect leaf motion.7 Even if a linear accelerator
existed that operated in a perfectly reproducible manner, there would still be the issue of inter-
and intra-fraction motion.8 Treatment plans are based on a snapshot of static anatomy prior to
treatment, and hence they introduce errors between Planned and Delivered Dose. The only
method for determining the cumulative effect of all these potential errors is to measure the entire
delivery sequence on the linear accelerator.
IMRT measurements can be classified into four levels: Level I measurements are taken using
film with or without ion chamber(s) placed in a static phantom; Level II measurements are taken
with the phantom moving to simulate intra-fraction motion; Level III measurements are taken
during each fraction of treatment, and delivered doses are reconstructed on the planning CT, and
Level IV measurements are dose measurements reconstructed on CT images acquired before or
during treatment. Ideally, Level IV measurements would be taken on a delivery system such as
the HI-ART (TomoTherapy, Inc., Madison, WI). At a minimum, Level I measurements should be
taken prior to treatment or within the first three treatment fractions.
Rebuttal
Certain external beam techniques (3D Conformal, Stereotactic Radiosurgery, etc ) and
brachytherapy procedures (High Dose Rate, Prostate, Cardiovascular, etc ) are routinely
commissioned prior to the treatment of patients. This is possible because the potential sources of
error in these techniques are limited in number and are quantifiable. The IMRT process is an
entirely different situation. There are hundreds to thousands of different sources of error
throughout the entire IMRT process. That is, there is an almost infinite list of possible
combinations of error. Therefore, it is simply impossible to commission the IMRT process in the
manner one uses with other treatment techniques.
64
The question that has confronted every medical physicist who has implemented IMRT is "How
much QA is enough?" The ACR and ASTRO have taken the position that verification must be
performed for each IMRT patient. This is to be done by irradiating a phantom that contains either
calibrated film or an equivalent measurement system to verify that the dose delivered is the dose
planned (i.e. Level I measurements). The purpose of this requirement is to ensure that the dose is
validated in at least one 2D plane on the actual treatment machine.
It is a mistake to substitute IMRT measurements entirely with a second MU calculation program.

Current IMRT MU programs use relatively simple algorithms to calculate the dose to one or
more points. The usefulness of these programs is limited to that of a "sanity check." Once Monte
Carlo calculational algorithms become available in IMRT planning systems, then it may become
appropriate to substitute some Level I and II measurements with calculations.
In conclusion, it is the responsibility of the medical physicist to promote the highest quality
medical services for patients. In the case of IMRT patients, it is necessary to perform individual
patient-specific testing to verify that the dose delivered is the dose planned.
AGAINST THE PROPOSITION: Scott Dube, M.S.
Opening Statement
The first commercial IMRT system was the NOMOS Peacock which included the Corvus
treatment planning computer and MIMiC multileaf collimator. This was an ingenious
advancement in radiation therapy whereby treatment plans were developed using inverse
planning software, and delivered with beamlets produced by a binary collimator operating
independently of the linac. The quality assurance (QA) for this novel technology was far more
rigorous compared with conventional radiation therapy. By necessity, it included chamber and
film measurements made in a phantom.9
Following the Peacock came multileaf collimator (MLC) IMRT. The treatment plan was
developed using a similar inverse planning process, but delivered by the MLC integrated with the
linac. It was natural that physicists responsible for MLC IMRT relied on the experience of
Peacock IMRT users for QA procedures. As a consequence, the tradition of chamber and film
measurements continued.
In 2000, the NCI convened a group of experts from AAPM and ASTRO to develop guidelines
and recommendations to help the radiation therapy community implement IMRT. 10 In their
report, these experts stated the need to commission the planning and delivery system as well as to
validate individual treatment plans. The commissioning is crucial because there are many sources
of potential error in the process, such as beam modeling of individual IMRT fields with complex
fluence maps, summation of individual fields to produce a composite distribution, transfer of
beam parameters from treatment planning system to linac, and delivery of the individual IMRT
fields by the linac. However, I question the value of validating individual plans. At our facility,
we have validated more than one hundred IMRT plans. Through this experience, as well as
private communication with many IMRT physicists, I have little evidence that validating
treatment plans is worthwhile after the commissioning process has been completed. The
additional effort spent for each patient is redundant with the initial testing.
65
My opinion is supported by the fact we do not validate individual treatment plans for other
complex procedures such as SRS or HDR. As with IMRT, proper commissioning will uncover
all potential errors, so the physicist can prevent them in the future.
The NCI report also states that monitor units (MUs) must be independently checked before each
patient's first treatment. This has been commonly done with chamber measurements in a
phantom. Fortunately for MLC IMRT programs, software has been developed that can validate
MUs by calculation.11 This is considerably faster than phantom measurements and yields similar
results. Reducing the validation effort frees up time for the physicist to evaluate the entire scope
of an IMRT treatment. For example, is the immobilization optimal for the patient? Are the dose
objectives consistent with recommendations found in the literature? Are there other beam
geometries to consider? Is there a need for TLD to assess skin dose?
In my opinion, the best use of a physicist's time would be to commission the IMRT
planning/delivery system, validate the MUs for individual plans, and play an active role in the
clinical development of IMRT.
Rebuttal
After considering the remarks of my colleague, I recognize three scenarios in which it would be
necessary to validate an individual IMRT plan before delivery.
1. The MU validation yields an unacceptable result. Whether using physical measurement or
independent software, there must be a threshold that triggers a further investigation by the
physicist.
2. The individual plan differs significantly from the types tested during the initial commissioning.
For example, assume the commissioning included a complete evaluation of a representative
prostate, nasopharynx, and brain case. If a new site is to be treated (such as breast) or if a new
technique is to be used (such as non-coplanar), then validation of that plan would be necessary.
Once this new type has been evaluated, it would not be necessary to validate subsequent plans of
the same type.
3. Some physicists decide not to perform an initial commissioning but instead validate each
individual IMRT plan as their program evolves. That can serve the same purpose of ensuring
correct treatments. However, there comes a point of diminishing returns when there is no value in
the redundant validation of similar plans. At that point, the validation of individual plans should
cease. The time otherwise spent on validating individual plans should be directed toward routine
quality assurance of the delivery system. It is critical to assure that the ongoing performance of
the MIMiC or MLC is correct.12 This will benefit all IMRT patients under treatment and not just
the new starts.
Keep in mind that the validation of individual IMRT plans is not a panacea. It offers no value in
addressing potential problems in the planning process (contouring, defining margins, assigning
dose objectives), daily delivery process (mechanical and radiological performance), or
anatomical registration (interfraction and intrafraction anatomical changes). The physicist must
address these issues by other activities.
REFERENCES
1. M. J. Zelefsky et al., "High-dose intensity modulated radiation therapy for prostate

cancer: early toxicity and biochemical outcome in 772 patients," Int. J. Radiat. Oncol., Biol.,
Phys. 53, 1111–1116 (2002).
66
2. C. Chao et al., "A prospective study of salivary function sparing in patients with head-
and-neck cancers receiving intensity-modulated or three-dimensional radiation therapy: initial
results," Int. J. Radiat. Oncol., Biol., Phys. 49, 907–916 (2001).
3. C. Chui et al., "Testing of dynamic multileaf collimation," Med. Phys. 23, 635–641
(1996).
4. T. LoSasso et al., "Physical and dosimetric aspects of a multileaf collimation system used
in the dynamic mode for implementing intensity modulated radiotherapy," Med. Phys. 25, 1919–
1927 (1998).
5. T. LoSasso et al., "Comprehensive quality assurance for the delivery of intensity
modulated radiotherapy with a multileaf collimator used in the dynamic mode," Med. Phys. 28,
2209–2219 (2001).
6. G. Ezzell et al., "The overshoot phenomenon in step-and-shoot IMRT delivery," J. Appl.
Clin. Med. Phys. 2, 138-148 (2001)
7. C. Mubata et al., "A quality assurance procedure for the Varian multi-leaf collimator,"
Phys. Med. Biol. 42, 423–431 (1997).
8. E. Huang et al., "Intrafraction prostate motion during IMRT for prostate cancer," Int. J.
9. D. A. Low et al., "Quality assurance of serial tomotherapy for head and neck patient
treatments," Int. J. Radiat. Oncol., Biol., Phys. 42, 681–692 (1998).
10. IMRT Collaborative Working Group, "Intensity-modulated radiotherapy: Current status
and issues of interest," Int. J. Radiat. Oncol., Biol., Phys. 51, 880–914 (2001).
11. J. H. Kung et al., "A monitor unit verification calculation in intensity modulated
radiotherapy as a dosimetry quality assurance," Med. Phys. 27, 2226–2230 (2000).
12. D. A. Low, "Quality assurance of intensity-modulated radiotherapy," Semin. Radiat.
Oncol. 12, 219–228 (2002).
67
2.2. Every patient receiving 3D or IMRT must have image-based

target localization prior to turning on the beam
Michael G. Herman and Douglas P. Rosenzweig

OVERVIEW
Some physicists contend that complex treatments such as 3D conformal therapy and IMRT
require tumor localization by an imaging technique such as electronic portal imaging each time
the patient is treated. Other physicists are willing to forego image-based target localization before
treatment, in part because such techniques are not reimbursable by 3rd-party carriers, and in part
because they believe such complexity is "precision overkill." This issue is debated in this month's
Point/Counterpoint.
Arguing for the Proposition is Michael G. Herman, Ph.D. Dr. Herman earned a Ph.D. in
experimental nuclear physics in 1986. He joined the Radiation Oncology Division at Johns
Hopkins in 1989 where he served on the faculty and in the capacities of Acting Chief and
Associate Director of Medical Physics. In 1998, he joined the Division of Radiation Oncology at
Mayo Clinic Rochester, where he is currently the Head of Physics, an assistant professor and a
member of the graduate faculty. He is board certified by the ABMP with a letter of equivalence
from the ABR. Dr. Herman serves actively in the AAPM, the ACMP, CAMPEP, the ABMP and
the ABR.
Arguing against the Proposition is Douglas Rosenzweig, Ph.D. Dr. Rosenzweig has been a
radiation therapy medical physicist for the last ten years. After receiving his Bachelors in physics
from the University of Wisconsin, he received his Masters and Doctorate degrees from the
University of Washington, working at its Nuclear Physics Laboratory. His dissertation involved
experiments probing pion physics. His medical physics experience began at the University of
Rochester, where he helped develop a prototype dynamically controlled collimator for
intracranial radiosurgery. He is certified by the ABMP, and has worked as a clinical medical
physicist in Milwaukee for the past five years.
FOR THE PROPOSITION: Michael G. Herman, Ph.D.
Opening Statement
The purpose of three dimensional conformal radiation therapy and intensity modulated radiation
therapy is to deliver high doses of radiation to the target volume, while minimizing dose to
normal surrounding tissues. Increased doses, sculpted specifically to the target, promise
improved control with fewer complications.
However, accurate target positioning within the treatment field is not guaranteed by 3D-
simulation, 3D treatment planning, three-point skin marks with lasers and weekly portal films.1
Each step adds uncertainty to the ability to deliver the prescribed dose to the target. Margins used
to establish the planning target volume (PTV) must quantitatively take these uncertainties into
68
account.2 The larger the uncertainties in patient positioning (setup error) and internal organ
motion, the larger the margins required. Magnitudes of target position variation have been
summarized,3 suggesting that margins implemented in the clinic could be inadequate to account
for these variations. It would seem that escalating doses or reducing margins should not be
routinely practiced until evidence of target localization for each fraction can be produced. The
lack of knowledge of where the target is within the field leads either to missing the target, or to
increased irradiation of normal tissue. This varies depending on the site being treated, but the
dosimetric and patient morbidity consequences are clear and we should do everything possible to
deliver the intended dose to the target.
Methods have been developed that allow the treatment team to assess and correct the target
position on a daily basis. These methods are efficient and effective in reducing target position
uncertainty, such that the PTV could be irradiated completely every day, and margins could even
be reduced. Electronic portal imaging devices (EPIDs) and ultrasound based systems are
examples of tools being used to reduce target position uncertainty. The clinical implementation
of EPIDs is detailed in AAPM TG58 report,4 and its application for daily PTV targeting has been
demonstrated.5,6 Trans-abdominal ultrasound localization has also been applied for daily target
verification.7 These approaches allow the treatment team to acquire statistically significant data
that permit patient-specific treatment improvements.8 It should be noted that whichever
technique is used, it must be properly commissioned and the process for use must be understood.
If we intend to increase dose to the target and reduce dose to surrounding tissues, it is essential
that we know where the target is through localization for each radiotherapy fraction. In most
cases, this can be done efficiently, but any extra effort required is well worth the knowledge that
the intended dose was delivered to the proper tissues within the patient.
Rebuttal
Dr. Rosenzweig correctly points out that we should not blindly recommend the daily use of
image-based target localization for every 3D and IMRT procedure. Nor should we blindly adopt
new and unproven treatment technology such as IMRT. The treatment team should determine the
frequency of imaging and weigh the costs and benefits as part of the implementation process of
3DCRT and IMRT. Only if the treatment team understands the patient-specific process of
simulation, planning and treatment delivery can the benefits of these treatments be realized.
ICRU report 62 describes the margins that must be considered to properly define a PTV, and
suggests that (internal and setup) margins are, in practice, combined subjectively based on the
experience of the treatment team. This suggestion reflects the absence of statistically significant
data to describe the margins. Furthermore, the report states that compromises in the prescription
dose or PTV due to this subjectivity can reduce the probability of treatment success. ICRU 62
does not provide enough guidance about PTV margins to ensure that we are delivering an
adequate treatment. It is the routine use of daily localization techniques that yields the patient-
specific and center-specific statistical data needed to define margins and deliver treatments
accurately. Perhaps not all conformal treatments require patient specific target localization, but
the team can only be certain of which treatments do need target localization if they have adequate
data.
Finally, the cost of daily imaging may not be insignificant, and the treatment team must develop
an effective protocol for the use of technology to improve the quality of treatment. Clinical
69
protocols are in place that use daily imaging to localize the target in a time efficient manner
without delivering any unplanned radiation dose to the patients.
I agree with Dr. Rosenzweig that we should have a flexible perspective on this topic. But this
perspective must be matched with the realization of what we are trying to accomplish with
radiotherapy. Using appropriate margins and localizing the target allows us to maximize the
benefit to the patient from highly conformal treatment technologies.
AGAINST THE PROPOSITION: Douglas Rosenzweig, Ph.D.
Opening Statement
Clinical practice is rapidly integrating 3D and IMRT (conformal) modalities, and clinicians will
soon rely on these techniques to treat even routine sites. Anticipating this expanding load, we
must be cautious in assessing the quality assurance requirements entailed, as there will be a broad
spectrum of situations with diverse clinical goals. The convenience of imaging devices makes it
tempting to rely on their daily use for control of treatment quality; yet such daily use has
important implications. Unquestionably, despite the increased convenience, daily imaging
introduces substantial operational costs. In addition, dosimetric issues arise regarding the volume
irradiated by daily orthogonal films. Alternative quality control strategies exist which may be
less burdensome and perhaps safer to nearby structures. Our profession has an obligation to
weigh the issue carefully as we enter this new era.
Conformal treatments can produce complex dose distributions. Undeniably, there is potential to
compromise therapy when such critically shaped distributions are geographically shifted. With
the diversity of treatment sites, though, we must clarify which classes are susceptible to danger,
and where we can proceed with more confidence. A blanket requirement of daily imaging for
every conformal treatment is premature, and could haunt us. Having recently seen our opinions
appropriated in rigid, unintended ways (e.g., the incorporation of TG-40 recommendations as
state-mandated QA requirements), we should learn from such examples.
Certainly, there are complex cases where daily imaging serves best to accurately and precisely
localize treatment. Because of interfraction organ motion, for example, daily electronic or
ultrasound imaging of conformal prostate therapy is likely the best targeting approach. Internal
fixation and precise and reproducible immobilization may offer comparable results, yet daily
imaging creates comparative efficiency and reduces patient discomfort. On the other hand, we
anticipate cases such as conformal palliative brain irradiation where daily imaging would be
excessive. Likewise, IMRT of the breast serves to produce homogeneity as opposed to dramatic
conformality, and given adequate immobilization methods we should here as well achieve
clinical goals without daily imaging. Along the spectrum of conformal cases, we will identify
situations where planners can safely define planning volumes based upon reasonable margins as
laid out in ICRU Report 62.9 These margins specifically account for expected random daily
geometric uncertainties. Grouping together all applications of conformal technology is unwise
when discussing quality assurance protocols.
Having adopted ICRU-62 planning margins, it is inconsistent to suppress random uncertainties

further at treatment time. Unfortunately, systematic setup errors remain possible, and daily
imaging does identify them. Alternatively, our literature suggests other strategies that can
efficiently address these problems early, 10 after which they are controllable through weekly
observation.11 Here, intelligent use of available tools reduces imaging requirements.
70
Physicists and oncologists must jointly decide the role and frequency of imaging. The costs and
efforts involved in daily imaging are substantial, and will be avoidable for some growing portion
of conformal treatments. We should be careful to have a flexible perspective on this topic, lest
our opinions be set in stone.
Rebuttal
Dr. Herman suggests that conformal treatments have goals of dose escalation and reduced
treatment volumes. These are specific goals of the initial application of this new technology,
which has focused on difficult cases where precise targeting is critical. However, clinicians and
vendors are working towards automation and efficiency that will allow broader application of the
technology. As integration progresses, conformal techniques will certainly be applied to more
routine cases where positioning accuracy and precision are already acceptable. For some of these
cases, less-frequent imaging strategies will be sufficient to achieve clinical goals with confidence.
It is our responsibility to understand the individual considerations of each class of treatment. Our
standard of practice should be flexible enough to account for the broadening landscape of cases.
Dr. Herman refers to several reports of successful daily imaging programs. It should be pointed
out that these cases are limited to imaging of the prostate, where extra effort is appropriate due to
dose escalation of a motile target near critical structures. Inferring from such cases that we can or
should expend similar efforts for all conformal treatments is not so straightforward. There are
many logistical issues that go into streamlining the imaging and decision-making processes
required to make daily verification suitably efficient. While these issues have been effectively
addressed in the case of prostate treatments, each treatment site has its own set of issues and
priorities. Regardless of one's efficiency, there will always be associated incremental costs. We
have an obligation to help control these costs at an appropriate level.
Unfortunately for all of us, our society has limited healthcare resources. We must strive to use
them appropriately, and must critically analyze additional demands on our clinical operations.
REFERENCES
1. R. K. Valicenti, J. M. Michalski, W. R. Bosch, R. Gerber, M. V. Graham, A. Cheng, J. A.

Purdy, and C. A. Perez, "Is weekly port filming adequate for verifying patient position in modern
radiation therapy?" Int. J. Radiat. Oncol., Biol., Phys. 30, 431–438 (1994).
2. M. van Herk, P. Remeijer, C. Rasch, and J. V. Lebesque, "The probability of correct
target dosage: Dose-population histograms for deriving treatment margins in radiotherapy," Int.
3. K. M. Langen and D. T. Jones, "Organ motion and its management," Int. J. Radiat.
Oncol., Biol., Phys. 50, 265–278 (2001).
4. M. G. Herman, J. M. Balter, D. A. Jaffray, K. P. McGee, P. Munro, S. Shalev, M. Van
Herk, and J. W. Wong, "Clinical use of electronic portal imaging: Report of AAPM Radiation
Therapy Committee Task Group 58," Med. Phys. 28, 712–737 (2001).
5. E. Vigneault, J. Pouliot, J. Laverdiere, J. Roy, and M. Dorion, "Electronic portal imaging
device detection of radioopaque markers for the evaluation of prostate position during
megavoltage irradiation: A clinical study," Int. J. Radiat. Oncol., Biol., Phys. 37, 205–212
(1997).
71
6. H. Alasti, M. P. Petric, C. N. Catton, and P. R. Warde, "Portal imaging for evaluation of

daily on-line setup errors and off-line organ motion during conformal irradiation of carcinoma of
the prostate," Int. J. Radiat. Oncol., Biol., Phys. 49, 869–884 (2001).
7. C. F. Serago, S. J. Chungbin, S. J. Buskirk, G. A. Ezzell, A. C. Collie, and S. A. Vora,
"Initial experience with ultrasound localization for positioning prostate cancer patients for
external beam radiotherapy," Int. J. Radiat. Oncol., Biol., Phys. 53, 1130–1138 (2002).
8. D. Yan, D. Lockman, D. Brabbins, L. Tyburski, and A. Martinez, "An off-line strategy
for constructing a patient-specific planning target volume in adaptive treatment process for
prostate cancer," Int. J. Radiat. Oncol., Biol., Phys. 48, 289–302 (2000).
9. ICRU Report 62: Prescribing, Recording and Reporting Photon Beam Therapy
(Supplement to ICRU Report 50) (ICRU, Bethesda, 1999).
10. J. C. J. de Boer and B. J. M. Heijmen, "A new approach to off-line setup corrections:
Combining safety with minimum workload," Med. Phys. 29, 1998–2012 (2002).
11. J. Hanley, M. A. Lumley, G. S. Mageras, J. Sun, M. J. Zelefsky, S. A. Leibel, Z. Fuks,
and G. J. Kutcher, "Measurement of patient positioning errors in three-dimensional conformal
radiotherapy of the prostate," Int. J. Radiat. Oncol., Biol., Phys. 37, 435–444 (1997).
72
2.3. Linear accelerators used for IMRT should be designed as small

field, high intensity, intermediate energy units
Tiru S. Subramanian and John P. Gibbons, Jr.
Reproduced from Medical Physics, Vol. 29, No. 11, pp. 2526–2528, November 2002
OVERVIEW
Linear accelerators are designed to provide large, uniform x-ray fields for use in conventional
therapy. But for IMRT, only a narrow fraction of the beam is used. Redesign of the linear
accelerator to IMRT specifications could enhance x-ray production efficiency, decrease cost, and
reduce photon and neutron dose outside of the useful beam. But this redesign limits the treatment
flexibility of the accelerator. The pros and cons of specially-designed linear accelerators for
IMRT are the focus of this Point/Counterpoint.
Arguing for the Proposition is T. S. (Mani) Subramanian, Ph.D. Dr. Subramanian has been a
practicing medical physicist in the San Francisco Bay Area for the past 15 years. Some of those
years were as an employee of the West Coast Cancer Foundation, and the others as a consulting
physicist. Before that he was a consultant physicist in South Florida and a hospital physicist in
Illinois. Dr. Subramanian obtained his M.Sc. and M.S. degrees from the University of Madras
and University of Detroit respectively, and a Ph.D. in experimental neutron physics at the
Crocker Nuclear Laboratory of the University of California, Davis. He has board certification in
Radiological Physics from the American Board of Radiology. He has been active with the Florida
Radiological Society and the San Francisco Bay Area Chapter of the AAPM.
Arguing against the Proposition is John P. Gibbons, Jr., Ph.D. Dr. Gibbons received his doctoral
degree in physics from the University of Tennessee–Knoxville in 1991. He was the first graduate
of the medical physics residency program of the University of Minnesota in 1993, and served on
the faculty there upon graduation. He is currently the chief physicist for Palmetto-Richland
Memorial Hospital in Columbia, South Carolina. He is board certified by the American Board of
Medical Physics and the American Board of Radiology. He also serves on several committees of
the AAPM and the ACMP, and is Past President and Board Representative of the South East
Chapter of the AAPM.
FOR THE PROPOSITION: Tiru S. Subramanian, Ph.D.
Opening Statement
In earlier years, radiation from cesium and cobalt units was used successfully (for the times) to
treat cancer. The limitations of this approach were high doses to superficial tissues, inadequate
depth doses, and lopsided tumor-to-normal tissue dose ratios for deep-seated cancers. As the
technology of linear accelerators evolved over time, supported by sophisticated imaging
modalities and dose calculation algorithms, techniques were evolved to provide improved, 3-D
conformal dose distributions for treatment of tumors.
73
In cancer management, both the physical (image-based, tumor-conforming dose distributions)

and the biological (tumor response and tissue complication relationships for tumors and normal
tissues) parameters play significant roles. While the physical parameters are readily understood
and improved by technology, the biological components are more difficult to quantify. In
addition, technological advances seem to warrant changes in protocols at a pace that is too fast
for the biological understanding to catch up with thorough well-designed clinical trials. Also, we
have barely crossed the threshold in the effective use of functional imaging modalities to assist in
treatment planning and delivery.
The conventional bremsstrahlung linear accelerator has been the mainstay in radiation oncology
for decades. IMRT techniques are implemented readily on these linacs with the help of MLCs.
The reliability of computer control and verification of linac and couch movements have improved
so that patient manipulation through couch movement now compensates for some of the
broadness of the treatment portal. Thus a smaller beam (and hence the smaller radiation head of
the linac) can provide many enhancements in the design of the linac, including a lower isocenter
and a greater patient clearance. Also, a smaller beam, or better yet, a scanning beam,1 means that
the thick flattening filter can be diminished or even eliminated, thereby enhancing the dose rate
available for IMRT. While the added dose rate may be too much to put to use now, the saying
goes that "if you build it, they will come!"
As for the optimum energy for IMRT, there is already a consensus building to focus on 6–8 MV.2
At this energy, the dosimetry characteristics are relatively well understood in heterogeneous
media. Also, 6–8 MV beams are less costly to shield. Currently, higher energies are also used for
IMRT for several reasons, including the desire to use all available energies, and the higher dose
rate at higher energies.
Rebuttal
Dr. Gibbons states "IMRT is an emerging technology that may become the treatment of choice
for perhaps 1/3 of radiotherapy patients." This probably is a factual statement, warranting the
question, "Is 1/3 of the patient load enough to plan for the optimized linac?" My contention is
that agreeing on a suitable low energy will make it worthwhile.
Dr. Gibbons strives to compare a dedicated IMRT (with usual electron capabilities included)
linac to a dedicated electron-only linac. That thought process has since delivered Mobetron to the
IOERT needs.
Dr. Gibbons acknowledges, "IMRT can certainly be delivered at lower energies using a
conventional linac," which supports the importance of energy optimization for IMRT. He also
recognizes "the inefficiency inherent in an MLC based collimating system," which supports the
need to redesign the linac head. This inefficiency occurs because current linacs are designed to
provide a 40×40 sq cm flat beam at isocenter. Never does IMRT require a field this large.
Dr. Gibbons fails to acknowledge the reliability of MLC-based linacs of today. There are many
facilities that have a single machine fitted with an MLC to meet the therapy needs of their
communities.
Finally, I wish to call on experts to start from a low energy setting and work upward to select the
optimum energy for IMRT, instead of the other way around. With this approach, the total body
74
dose from head leakage, internal scatter and neutrons will be minimized, especially in
comparison with enhanced energies above 10 MV.
AGAINST THE PROPOSITION: John P. Gibbons, Jr., Ph.D.
Opening Statement
IMRT is an emerging technology that may become the treatment of choice for perhaps 1/3 of
radiotherapy patients. However, it is not yet mature enough to drive manufacturing decisions.
The advantages of redesigning linear accelerators (linacs) to better accommodate IMRT are both
nebulous and insignificant, while the disadvantages are numerous. Until more data are available,
IMRT with conventional linacs is the more viable option.
First, let's look at the supposed advantages. It is argued that these machines would reduce cost.
However, it is more likely that the total cost will be larger, as departments would require both
conventional and IMRT linacs. Although "IMRT-only" linacs might be less expensive, they
would still cost more than an IMRT upgrade of a conventional linac. There would be additional
costs for vault construction, maintenance, etc., as well as additional personnel to operate the unit.
Indeed, if specialized linacs were more cost efficient, "electron-only" linacs (which would be
much less expensive than IMRT-only linacs) would already be widespread.
The argument that redesigned linacs could enhance x-ray production efficiency depends on the
specific configuration of the linac. For example, the increased output due to a reduction in
flattening foil thickness for smaller fields may be offset by the reduced beam intensity (~E–2) for
lower energies. Even if true, this advantage is not unique to a specialized IMRT linac: an
additional "IMRT filter" could be added to an existing linac. In many cases, however, it is the
collimating system rather than the linac, that limits IMRT dose rates, so improved x-ray
production is meaningless. Finally, there is no obvious clinical benefit to enhanced x-ray
production.
It is also stated that the photon and neutron dose outside the field would be reduced. The neutron
dose advantage is not unique: IMRT can certainly be delivered at lower energies using a
conventional linac. The clinical significance of a reduction of peripheral photon dose is not clear.
However, it is the inefficiency inherent in an MLC-based collimating system, rather than the
linac itself, that is primarily responsible for enhanced dose outside the field. A greater reduction
in photon leakage dose could be obtained by using physical attenuators with conventional linacs.
The fact that this technique is not widespread is itself evidence of clinical insignificance.
The proposed linacs also have several potential disadvantages. For example, some IMRT
treatment sites require large fields, which might not be achievable with an IMRT linac. It would
no longer be possible to treat IMRT and a conventional field concurrently without moving the
patient between fields. Lower energy treatments would increase dose to regions peripheral to the
target. Finally, it would be impossible to treat an IMRT patient on a conventional machine if a
specialized linac failed during the treatment course. This latter issue may necessitate the purchase
of a backup machine, which would add significantly to the cost.
Rebuttal
75
Dr. Subramanian summarizes my opening statement well when he says that "IMRT techniques
are implemented readily on these linacs ." There is no reason to change a working design,
especially if the change results in added cost and reduced flexibility.
My colleague claims that a redesigned IMRT linac offers the advantages of having a lower
isocenter and greater patient clearance. It has never been demonstrated, however, that there is a
need for either of these enhancements, given that existing systems can treat a wide range of
disease sites. Furthermore, it is not clear that either of these mutually-exclusive advantages truly
exists, because the reduction in treatment-head size would likely be perpendicular to, rather than
parallel with the beam direction.
In the same vein, there is no obvious clinical need for an enhanced dose rate machine. The
conjecture that some unknown clinical benefit may yet be found ("you build it, they will come")
is simply a radiation field of dreams. It is not clear that any significant enhancement could be
obtained, because existing collimation systems are the primary limitation in delivered dose rates.
In fact, depending on the linac design, dosimetric errors could increase due to communication
delays between collimation and linac control systems.3
Finally, Dr. Subramanian states that there is a consensus to utilize 6–8 MV. However, the study
he references concludes that lower energies require more MUs, beam segments and number of
fields, thereby increasing the overall treatment delivery time. Regardless, there is still no unique
advantage as both low and high energies are available on existing systems.
REFERENCES
1. S. Soderstrom et al., "Aspects of the optimum photon beam energy for radiation therapy,"
Acta Oncol. 38, 179–187 (1999).
2. A. Pirzkall et al., "The effect of beam energy and number of fields on photon-based
IMRT for deep-seated targets," Int. J. Radiat. Oncol., Biol., Phys. 53, 434–442 (2002).
3. G. Ezzell and S. Chungbin, "The overshoot phenomenon in step-and-shoot IMRT
delivery," J. Appl. Clin. Med. Phys. 2, 138–148 (2001).
76
2.4. Segmental MLC is superior to dynamic MLC for IMRT

delivery
Ping Xia and Joseph Y. Ting

OVERVIEW
Intensity modulated radiation therapy (IMRT) fields can be delivered with multileaf collimators
(MLCs) by either segmental or dynamic MLC methods. With segmental multileaf collimation
(SMLC) the leaves are stationary while the radiation beam is ON, which means that leaf
velocities are unimportant as far as intensity distributions are concerned. This is simpler than
dynamic multileaf collimation (DMLC), where the collimator leaves are moving while the beam
is ON and, hence, the velocity of the leaves is vitally important. This makes DMLC more
complicated to plan and deliver than SMLC. However, DMLC offers more degrees of freedom in
the design of intensity distributions. Some suggest that this extra versatility exhibited by DMLC
is insufficient to offset the simplicity of SMLC delivery of IMRT fields. This is the premise
debated in this month's Point/Counterpoint.
Arguing for the Proposition is Ping Xia, Ph.D. Dr. Xia is a clinical physicist and an Associate
Professor in the Department of Radiation Oncology, University of California, San Francisco
(UCSF). Dr. Xia received her Ph.D. in physics in 1993 from the University of Virginia,
Charlottesville. She obtained her training in radiation physics from UCSF in 1995–1997 and
joined the faculty there in 1997. She serves the AAPM as a member of Science Council, a
chapter representative on the Board of Directors, and a member of TG 119. Her main research
interests are treatment planning optimization and clinical implementation of emerging
technologies.
Arguing against the Proposition is Joseph Y. Ting, Ph.D. Dr. Ting received his Ph.D. degree from
Dartmouth College, Hanover, NH, in 1976 and, between 1974 and 1978, he pursued clinical
medical physics training at Allegheny Hospital, Pittsburgh, PA and at Thomas Jefferson
University Hospital, Philadelphia, PA. During his 30-plus years as a medical physicist, he has
held associate professorships at several leading academic medical centers and currently he is the
Chief Medical Physicist at MIMA Cancer Center, Melbourne, Florida. Dr. Ting has been
involved with IMRT patient treatments and pre-treatment IMRT physics quality assurance (QA)
issues since 1998. He is now involved with image-guided IMRT concerns and improvements in
treatment setup accuracy and reproducibility.
FOR THE PROPOSITION: Ping Xia, Ph.D.
Opening statement
Using a conventional multileaf collimator, segmental MLC (SMLC) for IMRT delivery is more
versatile than dynamic MLC (DMLC). SMLC can deliver IMRT plans as a simple extension to
the delivery method used for conventional three-dimensional conformal radiation therapy
77
(3DCRT). The report of the Intensity Modulated Radiation Therapy Collaborative Working
Group1 predicted in 2001 that ―widespread implementation of this form of IMRT is anticipated
during the next several years.‖ Although this prediction was based mostly on the fact that ―most
medical linear accelerator manufacturers are offering SMLC-IMRT capability,‖ the simplicity of
SMLC delivery is another reason that this method is well adapted for clinical use.
The pros and cons of SMLC and DMLC delivery methods have been debated since the early
clinical IMRT implementation a decade ago. 2 The main advantage of DMLC is that the
continuous leaf motion enables the delivered intensity profile to closely match the optimized
intensity profiles, particularly for the profiles produced by a beamlet-based optimization. The
advantages of the SMLC delivery method stem from its derivation from the conventional 3DCRT
delivery technique. It is, therefore, less demanding of the hardware control system of the linear
accelerator and permits more straightforward processes of verification and quality assurance. The
major disadvantage of the SMLC method is the prolonged treatment time when many segments
are used in the treatment plan, especially when each segment is recorded and verified during the
treatment. This disadvantage, however, has been eliminated by the recent development of direct
aperture-based optimization.3 It has been proven that the complex intensity profiles created by
beamlet-based optimization methods are not a requirement for highly conformal IMRT plans.
Several recent studies3,4 have demonstrated that direct aperture-based optimization can produce
plans that are equal to or even better than those from beamlet-based optimization, yet with
significant reductions in both the number of segments and the total number of monitor units
(MUs). Using breast treatment as an example, with up to six segments per tangential beam one
can obtain IMRT plans better than beamlet-based IMRT plans with improved plan quality, yet
with 60% fewer MUs. By contrast, dynamic MLC forced us to deliver even a simple field in a
complex way. For example, delivering 100 cGy to a 10×10 cm field with dynamic MLC
requires 180 to 2100 monitor units with window widths varying from 10 to 0.5 cm, respectively,
with ±6% variations in dose uniformity of the profile.5 For tumors involving respiratory motion,
the SMLC delivery method has even greater superiority over dynamic MLC as shown by Yu et
al.,6 who demonstrated that, when IMRT beams are delivered with dynamic MLC, the problem
of intrafraction motion causes large dose errors in every fraction.
In conclusion, the segmental MLC delivery method for IMRT is superior to the dynamic MLC
method, providing us with greater simplicity and flexibility for treating various types of cancers.
Furthermore, significant reduction in total monitor units will improve the accuracy of dosimetry,
reduce probability of radiation induced secondary cancer, and provide a safer environment.
AGAINST THE PROPOSITION: Joseph Y. Ting, Ph.D.
Opening statement
Long ago, I colored with a few crayons and put a few color patches onto a few numbered areas
on paper. It showed a picture. I was happy.
I developed an electronic tissue compensator in 1989. 7,8 It used a few jaw positions (field
segments) to generate field-in-field to replace physical wedges. It worked. I was happy.
Modern IMRT plans are composed of hundreds of beamlets and intensity levels. Why, then,
should one degrade this fine resolution by using coarse delivery methods?
78
IMRT delivery methods divide idealistic intensity maps from optimization routines into small
physical beamlets. Each beamlet has its unique intensity level. Dynamic or segmental deliveries
will result in identical delivered intensity maps if they both use the same number of beamlets,
segments, and intensities. If the segmental delivery were to deliver hundreds of intensity levels, it
would require many segments. It would take a long time setting each MLC segment, delivering
the radiation, and repeating the process. Therefore, most segmental deliveries are limited in the
number of segments and intensity levels. Some linear accelerators also impose a limit on the least
number of monitor units (e.g., 5) that they can deliver accurately.
With dynamic delivery, the radiation intensity is controlled by adjusting the duration that any
given location is exposed to radiation. It adjusts the gaps between pairs of opposing leaves while
the pairs move across the radiation field. A moving wide gap exposes the location to radiation
longer than a narrow gap.9 Computer-controlled servo motors operate the MLC leaves as they
move across the radiation field. This mode of operation does not thrust the leaf motors at full
speed in one direction, stop, then repeat this in the opposite direction. Only when one of the
leaves does not reach its target position for the required number of monitor units10 will the
radiation beam be held off until all leaves are at their predetermined positions. However, this
situation seldom occurs because the software configures the leaf motion according to hardware
and control system limitations.11,12
Concerns about dose delivery accuracy with ―dynamic MLC‖ are unfounded. Many investigators
have addressed IMRT QA with dynamic delivery. They have answered most of the questions
regarding QA techniques, such as varying intensities and sharp gradients across a measurement
volume, and electronic equilibrium issues. 10 Furthermore, most of these concerns are equally
applicable to segmental delivery because the fundamental composition of the intensity maps is
not different. Therefore, QA for dynamic delivery is not more difficult than for segmental
delivery.
In summary, if you are happy with a few intensity levels and segments, segmental delivery is for
you. It is like a summation of a few irregular fields. For example, there are clinics delivering
IMRT using just two segments. On the other hand, if you are an artiste worried about intensity
and spatial resolutions, and wish to preserve the integrity of an IMRT plan, and you want the best
for your patients, dynamic delivery is the most efficient and the best for your practice.
You can paint with four colors and ten patches or you can have a ―Rembrandt‖ with thousands of
colors and shades.
Rebuttal: Ping Xia, Ph.D.
I agree with Dr. Ting's analogy that in some aspects medicine is an art. A beautiful painting,
however, does not necessarily embrace hundreds or even thousands of different colors and
shades. A masterpiece of painting demands careful choices in colors and shades. Beginning
artists imitate masterpieces by dividing them into fine grids, confining their painting resolution to
the grid, or beamlet. Beamlet-based optimization in IMRT can be analogous to the beginner's
painting. It first divides the intensity pattern into grids, and then regroups a number of beamlets
into numerous patches (called leaf sequences). Even aided with hundreds of intensity levels, the
created intensity patterns are shadowy, making it difficult to envision a beautiful picture. The
recently developed aperture-based optimization eliminates the grid and directly optimizes the
shapes of radiation apertures and intensities, just as painting masters would do (in my opinion).
Aperture-based optimization allows us to create new colors when they are needed. In theory,
79
dynamic MLC can deliver segments created with aperture-based optimization, but the limited
mechanical speed of the MLCs may not be able to deliver these apertures. Certainly, one can use
dynamic MLC to deliver IMRT in the same fashion as segmental delivery. I believe that many
centers equipped with dynamic MLCs are doing so.
Rebuttal: Joseph Y. Ting, Ph.D.
A duck is a duck. Dr. Xia, we are saying the same thing. Simple fluence maps can be delivered
via either segmental or dynamic deliveries with ease. You state that it takes 180 – 2100 MUs to
deliver a ―flat‖ 100 cGy IMRT field with DMLC. However, I created a 10×10 field to deliver a
―flat‖ dose distribution at 10 cm depth using an electronic compensator with just 152 MU.
Besides, if one wishes to deliver a 10×10 open field, just set the jaws and deliver it. Why bother
with dynamic delivery?
Direct aperture optimization can work only if there are simple constraints and planning target
volumes. In breast treatments, inverse planning has no role. We use classical tangents with
electronic tissue compensators and dynamic delivery. MUs are similar to or less than
conventional wedge plans. While tangential breast fields can be delivered easily with relatively
few segments, for brain, head and neck, lung, abdomen, or prostate fields, segmental deliveries
are a gamble. Some times it works just fine, but it may not, and you will not know that unless you
compare both plans and delivery methods. I would stay with the tried and true.
Organ motion during IMRT delivery is not an issue. Multiple fractions and slow delivery will
average out these dose differences. Of course, there are methods to deal with organ motion
during treatment, such as gating and breath-hold.
An important message is lost. Dynamic delivery ensures the best reproduction of computed
fluence maps. Segmental deliveries may be simpler, faster, and easier to understand, but they
may not reproduce complex fluence maps properly. One may debate about the need for complex
fluence maps; but if you have complex fluence maps, you would need dynamic delivery to
reproduce them accurately.
In 1989, prior to MLC and inverse planning, I was delivering ―IMRT‖ fields using collimator
jaws with segmental deliveries. Segmental delivery works but you need to know when to use it.
REFERENCES
1. Intensity Modulated Radiation Therapy Collaborative Working Group, ―Intensity-

modulated radiotherapy: Current status and issues of interest,‖ Int. J. Radiat. Oncol., Biol., Phys.
51, 880–914 (2001).
2. C. S. Chui et al., ―Delivery of intensity-modulated radiation therapy with a conventional
multileaf collimator: Comparison of dynamic and segmental methods,‖ Med. Phys. 28, 2441–
2449 (2001).
3. D. M. Shepard et al., ―Direct aperture optimization: A turnkey solution for step-and-shoot
IMRT,‖ Med. Phys. 29, 1007–1018 (2002).
4. E. E. Ahunbay et al., ―Direct aperture optimization-based intensity-modulated
radiotherapy for whole breast irradiation,‖ Int. J. Radiat. Oncol., Biol., Phys. 67, 1248–1258
(2007).
80
5. R. Mohan et al., ―The impact of fluctuations in intensity patterns on the number of

monitor units and the quality and accuracy of intensity modulated radiotherapy,‖ Med. Phys. 27,
1226–1237 (2000).
6. C. X. Yu, D. A. Jaffray, and J. W. Wong, ―The effects of intra-fraction organ motion on
the delivery of dynamic intensity modulation,‖ Phys. Med. Biol. 43, 91–104 (1998).
7. E. S. Valentine, J. Y. Ting, and L. Davis, ―Sans wedge tangential breast treatment
technique,‖ Int. J. Radiat. Oncol., Biol., Phys. 21, 229 (1991).
8. H. E. Gaballa, G. Mitev, R. D. Zwicker, J. Y. Ting, and O. F. Tercilla, ―Tissue
compensation using dynamic collimation on a linear accelerator,‖ Int. J. Radiat. Oncol., Biol.,
Phys. 32, 723–731 (1995).
9. T. LoSasso, C. S. Chui, and C. C. Ling, ―Physical and dosimetric aspects of a multileaf
collimation system used in the dynamic mode for implementing intensity modulated
radiotherapy,‖ Med. Phys. 25, 1919–1927 (1998).
10. D. W. Litzenberg, J. M. Moran, and B. Fraass, ―Verification of dynamic and segmental
IMRT delivery by dynamic log file analysis,‖ J. Appl. Clin. Med. Phys. 3, 63–72 (2002).
11. P. Xia, C. F. Chuang, and L. J. Verhey, ―Communication and sampling rate limitations in
IMRT delivery with a dynamic multileaf collimator system,‖ Med. Phys. 29, 412–423 (2002).
12. D. W. Litzenberg, J. M. Moran, and B. Fraass, ―Incorporation of realistic delivery
limitations into dynamic MLC treatment delivery,‖ Med. Phys. 29, 810–820 (2002).
81
2.5. IMRT should not be administered at photon energies greater

than 10 MV
David S. Followill and Fridtjof Nüsslin

OVERVIEW
There is a significant increase in leakage radiation for an IMRT treatment compared with
conventional radiotherapy due to the increase in the number of monitor units required. The
resultant increase in whole body dose to the patient enhances the risk of radiation-induced cancer
and genetic effects. These risks are amplified if the IMRT treatments are delivered at photon
energies greater than about 10 MV due to neutron production, which increases the dose
equivalent to the patient. Some have suggested that this increased risk is not offset by any
advantages inherent in the use of higher photon energies and, hence, should be avoided by
administering IMRT treatments only at energies less than about 10 MV. This is the premise
Arguing for the Proposition is David S. Followill. Dr. Followill obtained his M.S. degree in
Nuclear Engineering from Texas A&M University, College Station, TX and his Ph.D. degree in
Biophysics from the University of Texas Health Science Center, Houston, TX. He completed
post-doctoral training in the Department of Radiation Physics at the M.D. Anderson Cancer
Center in Houston, where he has since spent his entire career. Currently he is an Associate
Professor and Associate Director at the Radiological Physics Center, University of Texas MD
Anderson Cancer Center. His major research interests have been radiation dosimetry and quality
assurance for radiotherapy, and neutron contamination around high-energy linear accelerators.
Arguing against the Proposition is Professor Fridtjof Nüsslin. Dr. Nüsslin received his Diploma
in Physics in 1966 and his Dr.rer.nat. in Physics and Physiology both from the University of
Heidelberg, Germany. He spent another two years for postdoctoral training in the Max-Planck-
Institute for Nuclear Physics in Heidelberg before joining the Department of Radiology at the
Medical School in Hanover, Germany for training in radiotherapy and radiology physics. In 1987
he was appointed Chair and Director of the Medical Physics Section of the Department for
Radiology at the University of Tübingen, Germany. In 2004 he became Professor Emeritus and
joined the Department of Radiation Therapy and Radiological Oncology of the University
Hospital of the Technical University, Munich, as Visiting Professor. There he continues to be
involved in research and training activities, mainly in radiation therapy, molecular imaging,
radiobiology, and radiation protection. Dr. Nüsslin continues to be active in many scientific and
professional organizations and, in 2006, he was elected Vice-President of the IOMP.
FOR THE PROPOSITION: David S. Followill, Ph.D.
Opening Statement
Clearly there is a great deal of data to support the statement that radiation can cause cancer.
Whenever radiation is delivered for beneficial reasons, the patient will still receive small doses of
82
scatter and leakage radiation to organs distant from the primary treatment site. For each organ,
the dose, no matter how small, comes with an associated risk of development of a secondary
cancer. The basis of this statement comes from the linear relationship between risk and radiation
dose derived from the A-bomb survivors.1,2 Some may argue that the risk is small and many of
our cancer patients may never live long enough to develop a second malignancy but, in reality,
more and more patients are living longer and survival rates for pediatric patients,3 where the risk
values may be 5 to 10 times that of an adult, are increasing. 4
Prior to the use of IMRT, the radiation oncology community considered the associated risks of
secondary cancers to be negligible when compared to the benefits of the radiation treatment.
However, with the implementation of IMRT where the number of monitor units has increased 3–
9 fold over that for conventional treatments, depending on the dose delivery technique and
planning system, the secondary doses and risks may no longer be considered negligible. 5,6 The
first IMRT treatments were performed using low (<10 MV) x-ray energies that did not include
any neutron contribution. Now, with the use of higher x-ray energies for IMRT treatments, the
secondary neutron dose, although small, becomes an important contributor to the total secondary
equivalent dose because of the neutron RBE.
The risks and benefits of each patient's radiotherapy treatment should be considered prior to
delivering the treatment. Any non-therapeutic dose of radiation that the patient receives
represents a potential risk. Data from Kry et al. has shown that IMRT treatments with energies of
10 MV or less from a specific make of accelerator, i.e. Siemens or Varian, result in less
secondary dose to the patient than treatments with higher energies. 6 This is mainly due to the
neutrons produced at the higher energies. The larger secondary doses reported at the higher x-ray
energies depended on the actual accelerator's photon energy spectrum, neutron RBE used,
number of segments, and total number of monitor units. Depending on the energy, accelerator,
and number of monitor units used, the risk of secondary fatal cancers can range from 2.2% to
5.1% for 15 and 18 MV energies, respectively.7 Taking into consideration the uncertainty in the
risk coefficients, the resulting risk from an IMRT treatment with 6 MV has been shown to be
statistically lower than for an IMRT treatment with 18 MV.8 IMRT treatments administered with
energies greater than 10 MV produce higher secondary doses and the higher the dose, the greater
the risk to our patients.
Therefore, if one wants to use a higher energy (>10 MV) for IMRT, the benefit of the resulting
treatment plan should have to clearly outweigh the increased risk of radiation-induced cancer
associated with the extra secondary radiation.
AGAINST THE PROPOSITION: Dr. Fridtjof Nüsslin
Opening statement
In order to debate the pros and cons of high energy (i.e., about 15 MV) photon IMRT, one needs
to remember some essentials of radiotherapy physics found in old textbooks: higher penetrative
quality, more pronounced build-up, lower skin dose, steeper dose gradients at the PTV margin,
better dose conformation to the PTV, and more effective dose sparing of normal tissue make high
energies the superior beam quality in many clinical situations. Even for head and neck tumors,
15 MV is preferred when the PTV is close to adjacent organs at risk. One disadvantage,
however, is the increased neutron leakage at higher energies, but this is small compared to scatter
from the useful beam. From a comparison of IMRT plans for head and neck treatments, it turns
83
out that for 10 and 15 MV the dose distributions are nearly identical. In a few cases, however,
about a 20% lower skin dose at higher energies will prevent the loss of patients' hair.
In a recent clinical study of IMRT lung treatment with 6 and 18 MV x-ray beams, Madani et al.9
demonstrated the minor role of energy selection in IMRT. However, what has not been evaluated
is the larger volume of normal tissues irradiated at lower doses, i.e., below 5 Gy. There are data
emphasizing low-dose radiation induced long-term risks of secondary malignancies10 and cardio-
vascular diseases.11 These kinds of long-term risks may strengthen the case for high energy
IMRT.
Pirzkall et al.12 studied the effect of beam energy and number of fields on IMRT for patients with
prostate carcinoma. They found a significant impact of energy selection on the dose distribution
in the region beyond a 1 cm margin around the target volume. Even when applying nine fields,
higher energies are to be preferred to achieve dose reduction in normal tissues and, hence, to
lower the risks for late effects.
In conclusion, as in conventional radiotherapy, there are good reasons to select higher energies in
IMRT techniques when treating thoracic and pelvic tumors, mainly in order to prevent late low
dose effects in normal tissues. In the head and neck region it may be prudent to apply lower
energies, except in a few situations where steep dose gradients at the PTV margin are essential,
or where skin dose reduction is expected to be clinically relevant. Finally, I would like to
emphasize the need for Monte-Carlo dose calculation models to achieve valid results when doing
this type of intercomparison study in IMRT treatment planning.
Although the impact of energy selection is rather small, I would prefer high energy IMRT when
dose reduction in normal tissue is an issue.
Rebuttal: David S. Followill, Ph.D.
The treatment decision for every patient is based on an analysis of the benefits versus the risks
for the particular treatment. Although a clinician may not rigorously document such an analysis,
the process still occurs as the patient's plan and prescription are approved. Clearly the energy to
be used to achieve an effective treatment is one of the most important decisions. Will the energy
selected achieve proper coverage of the PTV, reduce the dose to the surrounding normal tissues,
reduce the integral secondary dose near the target and, finally, leave the patient with an
―acceptable‖ risk of developing a secondary malignancy in distant organs away from the
treatment field? The patient's age, overall health, and life expectancy, are also important factors
to be considered in this decision-making process.
Perhaps a higher energy may deliver a plan that is dosimetrically better, but is the plan better for
the patient, especially a young patient (<45 years) with a long life expectancy? For some
patients, the risk of a secondary malignancy is irrelevant but, for young patients, the risk may
clearly outweigh any dosimetry benefit derived from using higher energies. The neutrons
generated by the higher energies and their associated RBEs clearly do not benefit the younger
patients. The differences between high and low energy IMRT are small and typically do not
affect the outcome of the therapy. The vast majority of radiation oncology centers continue to
deliver IMRT with 6 MV as their standard of care. The use of higher energies for IMRT may just
simply introduce another risk while not significantly improving the patient's treatment.
Selection of the beam energy for an IMRT treatment is only one of many factors that must be
considered when making treatment decisions and should not be the overriding factor in the
decision making process.
84
Rebuttal: Dr. Fridtjof Nüsslin
Dr. Followill stresses just a single argument, viz. the neutron contamination at higher energies
and its impact on secondary malignancies. Therefore, I conclude that we both agree about the
many favorable features of IMRT-treatment plans at higher energies (i.e., typically 15 MV, not
25 MV or more).
So what is in the neutron contamination argument? I agree with the statement that above 10 MV
the neutron production rate increases with photon energy. However, I disagree with nearly all the
speculative conclusions Dr. Followill presented and those in his Refs. 5,7,8. Is it scientifically
valid to apply the new radiation-weighting factor from the NCRP Report 116 for risk assessment
of developing fatal secondary malignancies despite the warnings not to apply the data of the
report to certain populations such as cancer patients? Dr. Followill is aware of the large
uncertainties of the risk assessments, but he and his colleagues in his Ref. 7 state ―until there
have been decades of evaluation time for patients treated with IMRT, the best estimate of the risk
from IMRT treatments is one calculated with these risk coefficients.‖ One can hardly believe in
any significant impact on energy selection in the limited range of 6 and 15 MV when a slightly
less than 0.3% difference in the calculated maximal absolute risk of fatal secondary malignancies
between 6 and 15 MV, for the same linac, has been calculated.7 The statement of a risk of fatal
secondary malignancies in the range of 2%–5% associated with leakage in an IMRT treatment is
not supported by facts and data from radiation biology and radiation epidemiology. Just one out
of many arguments omitted: about only 10% of all secondary malignancies following curative
radiotherapy are radiation induced, the remaining 80–90% are due to the increased life
expectancy.
Nevertheless, we are always challenged to minimize the radiation load to the normal tissue when
treating a patient. However, we have to set the priorities appropriately, i.e., increased tumor
control and lower complication probabilities, which clearly relate to the dose distribution, rank
higher than the speculative and quantitatively overestimated risks for secondary cancers.
Certainly manufacturers are challenged to carefully design their equipment before launching new
products with significantly higher leakage than well-established alternatives. I am referring to the
reported 4–5 times higher leakage of the CyberKnife in comparison to both the Leksell Gamma
Knife and linac-based IMRT systems without a proven gain in therapeutic efficiency.13
ACKNOWLEDGMENTS
Data provided by Markus Alber (head and neck study), Oliver Dohm (dosimetry), and Rainer
Schmidt (leakage radiation) is gratefully acknowledged.
REFERENCES
1. D. A. Pierce and D. L. Preston, ―Radiation-related cancer risks at low doses among

atomic bomb survivors,‖ Radiat. Res. 116, 178–186 (2000).
2. D. J. Brenner and R. K. Sachs, ―Estimating radiation-induced cancer risks at very low
doses: rationale for using a linear no-threshold approach,‖ Radiat. Environ. Biophys. 44, 253–
256 (2006).
3. L. Ries, C. Kosary, and B. Hankey, ―SEER cancer statistics review, 1973–1996,‖
Bethesda, MD, National Cancer Institute (1999).
85
4. International Commission on Radiological Protection: 1990 Recommendations. Annals

of the ICRP Publication 60 (Pergamon Press, Oxford, England, 1990).
5. D. Followill, P. Geis, and A. Boyer, ―Estimates of whole-body dose equivalent produced
by beam intensity modulated conformal therapy,‖ Int. J. Radiat. Oncol., Biol., Phys. 38, 667–672
(1997).
6. S. F. Kry et al., ―Out-of-field photon and neutron dose equivalents from step and shoot
intensity-modulated radiation therapy,‖ Int. J. Radiat. Oncol., Biol., Phys. 62, 1204–1216 (2005).
7. S. F. Kry et al., ―The calculated risk of fatal secondary malignancies from intensity-
modulated radiation therapy,‖ Int. J. Radiat. Oncol., Biol., Phys. 62, 1195–1203 (2005).
8. S. F. Kry et al., ―Uncertainty analysis for calculated risk of secondary fatal malignancies
from radiotherapy treatments including IMRT,‖ Int. J. Radiat. Oncol., Biol., Phys. (accepted).
9. I. Madani et al., ―Comparison of 6 MV and 18 MV photons for IMRT treatment of lung
cancer,‖ Radiat. Oncol. 82, 63–69 (2007).
10. J. D. Boyce et al., ―Cancer in the contralateral breast after radiotherapy for breast
cancer,‖ N. Engl. J. Med. 326, 781–785 (1992).
11. Early Breast Cancer Trialists' Collaborative Group, ―Favourable and unfavourable effects
on long-term survival of radiotherapy for early breast cancer: an overview of the randomized
trials,‖ Lancet 355, 1757–1770 (2000).
12. A. Pirzkall et al., ―The effect of beam energy and number of fields on photon-based
IMRT for deep-seated targets,‖ Int. J. Radiat. Oncol., Biol., Phys. 53, 434–442 (2002).
13. P. L. Petti, C. F. Chuang, V. Smith, and D. A. Larson, ―Peripheral doses in CyberKnife
radiosurgery,‖ Med. Phys. 33, 1770–1779 (2006).
86
2.6. New radiation therapy rooms should be designed to

accommodate treatments of the future such as IMRT and
tomotherapy
William G. Van de Riet and Richard G. Lane

OVERVIEW
Radiation therapy is experiencing major changes with the advent of techniques such as conformal
therapy, IMRT, tomotherapy and HDR brachytherapy. These techniques permit refined patterns
of dose delivery including dose escalation and field shaping to provide improved therapeutic
ratios between tumor and normal tissues. They also impose new design limitations and flexibility
requirements on treatment rooms that can increase their cost and complexity. Whether these
limitations and greater flexibility, with their attendant costs, should be accommodated through a
more modular approach to therapy room design is the subject of this Point/Counterpoint issue.
Arguing for the Proposition is William Van de Riet, Ph.D. Dr. Van de Riet received his Ph.D.
from the University of Kansas and is currently practicing Radiological Physics at Bloomington
Hospital in Bloomington, Indiana. He is certified by the American Board of Radiology and has
served as an examiner for that organization on several occasions. He has also served on several
committees of the AAPM and was on the Radiation Advisory Board of the Michigan Department
of Health for many years. He has been involved in the design of more than 30 radiation treatment
rooms during his career.
Arguing against the Proposition is Richard G. Lane, Ph.D. Dr. Lane received his doctoral degree
from the University of California-Los Angeles in 1970. He has held faculty positions at the
University of Wisconsin, the University of New Mexico, and the University of Texas Medical
Branch at Galveston. He is now a Professor and Group Leader of the External Beam Therapy
Services Group in the Department of Radiation Physics at University of Texas M. D. Anderson
Cancer Center. He is board certified by the American Board of Radiology in Radiological
Physics and by the American Board of Medical Physics in Radiation Oncology Physics. He has
served on several AAPM committees, and has been President of three chapters of the AAPM. He
also serves on committees of the ACMP, the ACR, and is presently the Chair of the CAMPEP
Residency Education Program Review Committee. Dr. Lane has published over 50 papers in
peer-reviewed journals.
FOR THE PROPOSITION: William Van de Riet, Ph.D.
Opening Statement
New radiation therapy treatment rooms should be designed to accommodate treatments of the
future such as IMRT and tomotherapy. This statement reflects my opinion that the evolution of
treatment units makes this easier, the design considerations are not that extensive or costly, and
new technology eventually reaches most radiation treatment facilities.
87
Megavoltage treatment units have changed dramatically since the 1950s and 1960s. The physical
techniques to produce high energy photons has changed, treatment units have shifted from fixed
to rotational, the maximum photon energy available has increased, and the rotational isocenter
distance has changed. Those of us who practiced during these changes can relate the frustrations
we experienced when we had to abandon or dramatically remodel treatment rooms that were too
small or drastically undershielded. Fortunately, this situation has settled down to a large extent.
Today, most treatment rooms are designed for a dual energy electron linear accelerator with 100
cm isocenter distance, maximum energy of 15–23 MV, and multileaf collimation. Most of us do
not foresee big changes in these aspects of treatment units in the near future, and a room designed
today for conventional treatments should be functional for the next 10–20 years. Because these
aspects have stabilized and computer technology has advanced so rapidly, manufacturers and
large medical centers have concentrated since the mid 1990s on refining treatment techniques.
The major design considerations for IMRT and tomotherapy are increased leakage radiation and
additional cooling. IMRT using a linac with MLC, serial tomotherapy, or helical tomotherapy
may increase the leakage workloads by factors of 5, 10 and 20 respectively. As a result, the
secondary barrier thicknesses have to be increased by 8–12 inches of concrete (helical
tomotherapy units use lower energies of 6–10 MV and have additional head shielding). In
addition, for IMRT and serial tomography, the increased neutron fluence can be dealt with by
designing a longer maze, turning the entrance door 90 degrees, or using a 25 MV rather than a 15
MV neutron door. These new modalities also increase the cooling requirements of the room by
25 to 30% compared to conventional workloads. Since the labor cost for constructing forms is
about the same for a 36 or 48-inch thick wall, the additional cost would only be for material and
possibly space. The heating aspect could be taken care of by installing a cooling unit of larger
capacity. All of these considerations can be accomplished at substantially reduced cost and with
less disruption of service if they are added at the time of initial construction.
New technologies eventually filter down to the community hospital. Hence, new treatment rooms
should be designed to accommodate these new modalities.
Rebuttal
Dr. Lane and I agree that large institutions with multiple treatment units will probably dedicate
treatment rooms for IMRT. These large institutions will take into account the additional
secondary radiation associated with IMRT when they design the dedicated rooms. He argues,
however, that no additional design considerations need to be made for conventional treatment
rooms housing dual energy photon beam linear accelerators. His reasons are that IMRT
techniques will be performed with low energy photon beams, and that the slow throughput will
result in the same overall workload as conventional treatments.
Although serial and helical tomography are performed using low energy photons, this is not
necessarily the situation for dynamic or static multileaf IMRT. Multileaf collimators are already
designed to accommodate high energy photons, and most lesions treated with high energy beams
using conventional techniques will likely be treated with high energy beams using IMRT
techniques.
I agree that patient throughput for IMRT is slower than for conventional treatments at the present
time. But as the output of linear accelerators increases, and computer control of these units
becomes faster, the throughput will increase. Just a few years ago, IMRT was performed with
88
cast blocks, compensating filters and outputs of 300 cGy/minute or less. A few years from now,
faster throughput for IMRT will likely yield substantial increases in secondary workloads.
Taking the time to consider "what if" now may pay off in the future. My argument remains that it
is much cheaper to provide some "overkill" now than to face large remodeling costs in the future.
AGAINST THE PROPOSITION: Richard G. Lane, Ph.D.
Opening Statement
Preliminary outcomes seem to predict that IMRT is the treatment of choice for about one third of
patients treated with external beam radiation therapy. The great concern over accommodating
IMRT treatments is that they use radiation inefficiently. The IMRT tumor dose is not much
greater than that for conventional treatment, but the number of monitor units (MUs) required to
deliver this dose is greater by an order of magnitude for IMRT treatments delivered by sequential
or helical tomotherapy. Dynamic or static multileaf IMRT techniques use 2–5 times as many
MUs. This requirement greatly increases the leakage radiation and, potentially, the neutron
production for each patient treatment. How should one accommodate this aspect of IMRT in the
design of new treatment vaults?
Large facilities with multiple treatment units may have dedicated sequential or helical
tomotherapy IMRT units in vaults designed specifically for IMRT. However, most facilities will
alternate between conventional and IMRT treatments using a dual-use, dual-photon energy,
linear accelerator providing low energy photons of about 6 MV and high-energy photons of 15–
25 MV. It is my position that there is no need to make special accommodations for IMRT in the
room design for a dual-use treatment unit.
The use (U) factor and workload (W) for conventional treatments using the high-energy beam
will dominate the shielding calculations for both the primary and secondary barriers. There is
little or no difference between IMRT treatment plans that use low-energy photons and those that
use high-energy photons. No accommodation needs to be made for increased neutron production
because IMRT treatments will be delivered with low energy photons.
It has also been shown that IMRT treatment times are three to four times longer than those for
conventional treatments. No accommodation needs to be made for increased leakage in the
shielding calculation, however, because the reduction in patient throughput (workload)
corresponds almost exactly to the increase in leakage radiation per patient. And the secondary
barrier, shielded for conventional, high-energy photon treatments, will easily provide sufficient
protection for a small increase in total leakage of low energy photons.
So, there is no need to make expensive room design accommodations for implementing IMRT
techniques in radiation therapy.
Rebuttal
The argument presented is that new vaults should be built to accommodate IMRT and
tomotherapy because a costly accommodation will be made at some time and that current
construction costs are lower than future costs. However, this accommodation will probably never
be required.
89
IMRT is an evolving technology with solutions to several problems already being implemented.
For example, serial tomotherapy exhibits some dose uncertainty between treatment slices. Helical
tomotherapy addresses this issue. Also, helical tomotherapy incorporates a CT imaging system
designed to avoid a geometric miss by providing daily tumor volume localization. In addition,
helical tomotherapy incorporates increased head shielding to reduce leakage radiation.
Dynamic multileaf collimation (DMLC) IMRT also addresses the issue of dose uncertainty.
However, the multileaf collimator (MLC) was originally designed to be a block replacement
device for conventional treatments that average about 4 MLC positions per patient. DMLC IMRT
treatments may require over 100 MLC positions per patient. A new generation of MLC is
designed to address the issue of MLC reliability under these conditions.
Regulations limit leakage radiation to approximately 0.1% of a conventional treatment tumor

dose. This is not only to protect individuals outside the vault but also to limit the total body
exposure of the patient. With IMRT leakage workloads that are ten times that of conventional
treatments, the exposure of the patient increases to about one percent of the tumor dose. Total
body doses approaching 100 cGy may well have an adverse effect on long-term treatment
outcomes. Increased leakage radiation associated with IMRT is a major issue that will be
addressed by increased head shielding or by changes in IMRT treatment technique. For example,
there are IMRT techniques under investigation that are quite efficient in the use of radiation.
These include step-and-shoot MLC IMRT and physical attenuator IMRT.
Most radiation oncology facilities will offer IMRT within three to four years. Increased patient
awareness of potential improved outcomes combined with increased reimbursement will ensure
that all patients have access to this treatment modality. Major problems associated with IMRT
will be resolved through evolution of the technique and through modification of the equipment
by the vendors.
There is no need to make expensive room design accommodations for implementing IMRT
treatment techniques.
90
2.7. IMRT may be used to excess because of its higher

reimbursement from medicare
Bhudatt R. Paliwal and Ivan A. Brezovich

OVERVIEW
Intensity-Modulated Radiation Therapy (IMRT) is reimbursed by Medicare at a significantly

higher rate than conventional radiotherapy. Some physicists are convinced that this differential
reimbursement expands IMRT applications well beyond those justified by improved patient care.
Others believe that IMRT improves patient care in almost all patients eligible for curative
radiation therapy. This controversy is the topic of this month's Point/Counterpoint.
Arguing for the Proposition is Bhudatt Paliwal, Ph.D. Dr. Paliwal is Professor of Human
Oncology and Medical Physics at the University of Wisconsin in Madison, WI. He is also the
Director of Medical Physics in the radiation oncology clinic, and has served in this capacity for
twenty-five years. His major areas of interest are image-guided therapeutics, IMRT,
brachytherapy and tomotherapy.
Arguing against the Proposition is Ivan Brezovich, Ph.D. Dr. Brezovich is Professor of Radiation
Physics at the University of Alabama at Birmingham, AL. He is the Director of the Radiation
Physics Division in the Department of Radiation Oncology, and has served in this capacity for
almost twenty years. His major areas of interest are IMRT, respiration gated dose delivery, and
brachytherapy.
FOR THE PROPOSITION: Bhudatt Paliwal, Ph.D.
Opening Statement
Intensity-modulated radiation therapy (IMRT) is a state-of-the-art cancer treatment method that

delivers high doses of radiation directly to cancer cells in a very conformal way, much more
precisely than is possible with conventional radiotherapy. This technique can deliver higher
radiation doses directly to the target while sparing more of the surrounding healthy tissue. IMRT
requires expensive equipment and extra labor on the part of the physician, physicist, dosimetrist
and radiotherapist, all of whom are essential members of the IMRT treatment team. IMRT is a
relatively complex and ex pensive process compared with conventional radiation therapy. The
extra cost is paid by third-party carriers, state and federal healthcare agencies, and patients.
IMRT is not a fundamental breakthrough that changes the tumoricidal effect of radiation. It is
merely a refinement 1 in the delivery process. It impacts patient care for a subset of the cancer
patient population. The efficacy of IMRT should be assessed by the probability of local control.
If there were no change in local control, the use of IMRT would be a waste of resources. Even
though the enthusiasm for gains in outcome is high, the actual data are sparse to substantiate the
claims of IMRT as a more effective treatment course.
91
There are potential negative aspects of IMRT. Precision radiotherapy by IMRT should be
delivered only to targets that can be precisely defined. Many tumors are not imaged well enough
to achieve a high degree of precision. An arbitrarily defined target volume has the potential for
under-coverage of tumors or over-treatment of normal tissue. IMRT results in a significant
increase in total body dose from increased treatment time and leakage radiation.
Because of a higher reimbursement by CMS (Medicare) for IMRT treatments, entrepreneurs,
consciously or unconsciously, may be more likely to over-utilize IMRT to enhance the financial
gain. Even small clinics with limited resources are pushing IMRT in order to compete in an
aggressive market.
Unnecessary usage of IMRT can occur at several clinical sites. IMRT is particularly unjustified if
there is no particular need to avoid normal tissues. In the case of the vocal cords, for example,
tumors have been treated very successfully with standard therapy. Using IMRT would be
overkill. Other sites, such as lung, brain and liver where there is a high risk of treatment failure,
may yield minimal benefits for IMRT.
In general, a busy radiotherapy clinic treats about half of its patients with curative intent. It is
estimated that about 20 percent of curative patients would derive some benefit from IMRT.
Treatment of a significantly higher fraction would be an over utilization of IMRT.
Rebuttal
Dr. Brezovich lists the medical physics developments over the last century to support his
position, but does not give any rational basis for making an informed decision for appropriate use
of IMRT. This is precisely the problem. We have developed a highly complex, computer
controlled technology without providing a decision-making process to determine when and
where to use it. Dr. Brezovich also states that "In its current state of development, IMRT is very
labor intensive. Excessive use of IMRT occurs only when corners are cut at the expense of
quality, in order to generate unearned profits." This is exactly my point. Do I need to argue
further?
Precision dose delivery is a lofty objective. Physicists, chemists and pharmacists all strive for a
very high level of accuracy in measuring and calculating quantities. They can measure in units of
nanometers or millionths of a gram. It would be foolish, however, to measure height and weight
of patients to such an accuracy because there is no treatment that demands such accuracy.
Similarly, it is absurd to postulate that all radiotherapy treatments require the highest order of
accuracy that is technologically achievable and fiscally affordable.
The zeal to achieve high accuracy should not be the primary determining factor in the use of
IMRT treatment for a patient. The degree of accuracy achieved should be that necessary to
achieve the maximum possible therapeutic gain. But the science of optimizing therapeutic gain
lags significantly behind the capability to deliver IMRT. It therefore permits over-utilization and,
in some cases, abuse with the intent of achieving greater financial gains. Obviously, physicians,
healthcare organizations and vendors recognize the higher reimbursement for IMRT procedures.
They are vulnerable to the pressure of marketing forces and reimbursement increases. It is
doubtful that the present trend of generous reimbursement will continue if there is excessive use
without clear demonstration of improved outcome.
A major drawback of IMRT is its inability to verify treatment fields in vivo. Even though the
science of calculating and planning treatments has achieved a high degree of accuracy, the
92
science of precision positioning of organs and tissues in a virtual space defined by the treatment
planning system has a long way to go. Patients breath, wiggle, scratch and move while on the
treatment couch. The tight margins that are often advocated for IMRT in order to escalate tumor
dose lead to greater uncertainty in peripheral doses. This can produce a significant mismatch
between the precisely calculated tumor dose and the actual delivered dose. As a consequence, it
is unwise to propose IMRT as a generalized prescription for all treatment sites.
It is important to quantify the improvement in outcomes from IMRT. The use of this technology
can only be justified through definitive therapeutic gains. Fifty percent of radiotherapy patients
are treated with a palliative intent, and it is difficult to see how these patients would benefit from
IMRT. In about one-third of the curative population, conventional treatment schemes are very
effective in controlling local disease. And if one takes motion uncertainty into account, only
about 20% of the total radiotherapy patient pool can be justifiably treated with IMRT. In the
current environment, a higher usage is unjustified overuse.
AGAINST THE PROPOSITION: Ivan Brezovich, Ph.D.
Opening Statement
Precise dose delivery has been a major objective of medical physics research ever since ionizing
radiation was introduced into the clinic more than a century ago. The recent transition to the TG-
51 calibration protocol exemplifies the continuing progress in quantifying radiation.
Improvements in spatial dose definition include the replacement of fixed circular cones by
continuously adjustable collimators, and custom Cerrobend blocks replacing hand-placed blocks.
Photons from high-energy accelerators spare the skin when deep tumors are treated, whereas
electrons avoid damage to deep structures when superficial lesions are targeted. The introduction
of multileaf collimators in conjunction with high-speed computers made conformal 3D therapy
practical. In that context, IMRT is just another step toward better dose localization, likely to be
followed by a wider use of protons and other particles such as mesons that allow even more
precise targeting.
The ability of IMRT to deliver sophisticated dose patterns with great accuracy has been
established in numerous phantom experiments and Monte Carlo simulations. 2,3 Its ability to
deliver therapeutic doses to tumors in the head and neck, while sparing the parotid, has also been
proven.4 Similar performance has been observed in the retreatment of vertebral bone metastases5
and prostate cancer.6 In tangential breast therapy, IMRT improved dose uniformity.7,8 There is
also evidence that the superior dose patterns translate into improved local control and quality of
life.5,6
In its current state of development, IMRT is very labor intensive. Complicated cases often
require over 10 hours of a medical physicist's time for planning and phantom verification before a
patient can be safely treated. Except for extensive trauma surgery, few procedures require as
many high-intensity hours of medical specialists. Medicare is well aware of the critical nature of
IMRT, and provides reimbursement commensurate with the cost of equipment and specialist
time. Excessive use of IMRT occurs only when corners are cut at the expense of quality, in order
to generate unearned profits.
Skeptics argue that IMRT diverts scarce money from areas where it could be more beneficial.
Therefore, IMRT should be used only rarely, if at all. However, the equitable allocation of
93
government funds is the responsibility of Congress, and medical physicists must function within
the framework laid down by our elected representatives. Medical physicists should strive to use
IMRT for every patient who can benefit from it. Furthermore, costs will drop as IMRT matures.
Even today, IMRT is often delivered as quickly as standard treatment, because it eliminates the
need for time-consuming changing of blocks and wedges.8 Inverse planning for IMRT is
amenable to further automation, and may eventually take less time than conventional planning.
Radiation oncologists may become tired of looking at consistently near-perfect treatment plans,
and turn over some of the simpler cases to physician extenders, at great savings. The potential of
IMRT to provide better patient care at a lower cost greatly outweighs any eventual short-term
waste caused by excessive use.
Rebuttal
I share Bhudatt's concern that, at least in some cases, IMRT may hurt patients more than help
them. An article in the New England Journal of Medicine9 shows that—against common
expectations—cancer deaths rise with increasing cost of treatment. The article points out,
however, that the higher treatment costs at some cancer centers in Florida are correlated with
fewer medical physics hours spent per patient. So the elevated cancer death rate could be caused
not by the higher cost, but by reduced treatment quality at some of the more expensive facilities.
Precise dose delivery afforded by IMRT does not compel radiation oncologists to cut margins
beyond safe tumor coverage. It does allow them to exclude tissues known to be disease free, to
the immediate benefit of patients. As knowledge concerning tumor extent progresses, margins
can be further narrowed. Since normal tissue tolerance is often the dose-limiting factor, dose
escalation to small, poorly oxygenated regions of the tumor may increase tumor control rate and
survival. Unless we abandon the paradigm that geographically-precise dose delivery is desirable,
IMRT spells progress.
Compared with overall healthcare expenditures, IMRT is modestly priced. Distributed over all
patients treated during the service life of an accelerator, the equipment cost per patient is well
below $1,000. Personnel costs will also drop with increased automation, potentially transforming
IMRT into a cost-cutting technology that eliminates the expensive salvage treatments subsequent
to failure of initial treatment. Thus, the current high cost to Medicare is simply an initial
investment rather than a permanent cost escalation. For patients to derive optimal benefits from
IMRT, medical physicists need to become more politically involved, to assure that money spent
on behalf of patients is used for their benefit.
REFERENCES
1. E. Glatstein, "Intensity-modulated radiation therapy: The inverse, the converse, and the
perverse," Semin Radiat. Oncol. 12, 272–281 (2002).
2. N. Agazaryan, T. D. Solberg, and J. J. DeMarco, "Patient specific quality assurance for
the delivery of intensity modulated radiotherapy," J. Appl. Clin. Med. Phys. 4, 40–50 (2003).
3. A. Leal et al., "Routine IMRT verification by means of an automated Monte Carlo
simulation system," Int. J. Radiat. Oncol., Biol., Phys. 56, 58–68 (2003).
4. N. Lee et al., "Intensity-modulated radiotherapy in the treatment of nasopharyngeal
carcinoma: an update of the UCSF experience," Int. J. Radiat. Oncol., Biol., Phys. 53, 12–22
(2002).
94
5. S. Milker-Zablel et al., "Clinical results of retreatment of vertebral bone metastases by

stereotactic conformal radiotherapy and intensity-modulated radiation therapy," Int. J. Radiat.
Oncol., Biol., Phys. 55, 162–167 (2003).
6. M. J. Zelefsky, Z. Fuks, and S. A. Leibel, "Intensity-modulated radiation therapy for
prostate cancer," Semin Radiat. Oncol. 12, 229–237 (2002).
7. E. M. Donovan et al., "Dose-position and dose-volume histogram analysis of standard
wedged and intensity modulated treatments in breast radiotherapy," Br. J. Radiol. 75, 967–973
(2002).
8. F. A. Vicini et al., "Optimizing breast cancer treatment efficacy with intensity-modulated
radiotherapy," Int. J. Radiat. Oncol., Biol., Phys. 54, 1336–1344 (2002).
9. J. M. Mitchell and J. H. Sunshine, "Consequences of physicians' ownership of health care
facilities-joint ventures in radiation therapy," N. Engl. J. Med. 327, 1497–1504 (1992).
95
2.8 Future developments in external beam radiotherapy will be

unlikely to significantly improve treatment outcomes over those
currently achieved with 3D-conformal and IMRT treatments
Robert J. Schulz and Dirk L. J. Verellen
OVERVIEW
Over the past decade, 3D conformal therapy and, especially, IMRT, have been heralded as major
advances in external beam radiotherapy that are likely to significantly improve outcome.
Technology has not stopped there, however, and we are faced with many new developments such
as image-guided radiotherapy (IGRT), respiratory gating, tomotherapy, proton and heavy ion
therapy, etc., all of which might improve outcomes even further. Unfortunately, each of these
new technologies increases the cost of radiotherapy, so it is important to address cost/benefit
concerns: will the improvements in outcome be significant enough to justify the increased costs?
This is the topic of this month's Point/Counterpoint debate.
Arguing for the Proposition is Robert J. Schulz, Ph.D. Dr. Schulz is a charter member and
Fellow of the AAPM, Fellow of the ACR, and Diplomate of the ABR. His professional career
began at Memorial Sloan-Kettering (1952–1956), developed further at the Albert Einstein
College of Medicine (1956–1970), and concluded at Yale University (1970–1992) from which he
retired as Emeritus Professor. His major contributions have been in radiation dosimetry having
chaired the SCRAD and TG-21 committees, and twice been a recipient of Farrington Daniels
Awards. His retirement to northern Vermont, and close friendship and collaboration with A.
Robert Kagan, M.D., has broadened his perspective on radiation therapy to include
considerations of cost versus benefit, over-diagnosis and over-treatment, the quality of clinical
reports, and the impact of new technologies on outcomes.
Arguing against the Proposition is Dirk Verellen, Ph.D. Dr. Verellen is Professor, Vrije
Universiteit Brussel (VUB), and director of the Medical Physics Group in the Department of
Radiotherapy, Universitair Ziekenhuis, Brussels, Belgium. He received his M.Sc. in solid-state
physics from the University of Antwerp, Belgium, and his Ph.D. in medical sciences from the
VUB. In addition to his educational duties at the VUB in Medical Physics, he is also guest
professor at the Europese Hogeschool, Brussels, in the training program for Radiation
Technologists and course director for Image-guided Radiation Therapy at the ESTRO European
School of Radiotherapy. He is author of over 40 peer-reviewed scientific papers and editor or
contributing author of several books. His main interest is in the clinical implementation of
conformal radiotherapy and image-guidance. He serves on the Board of Editors of Medical
Physics, and chairs the Working Group on New Technologies for the Organization of European
Cancer Institutes. He is a member of the Nederlandse Commissie voor Stralingsdosimetrie sub-
committee on Guidelines for Stereotactic Treatments, the AAPM Task Group 101 on Stereotactic
Body Radiosurgery, and the Belgian Quality Audit Programme for Radiotherapy.
FOR THE PROPOSITION: Robert J. Schulz, Ph.D.

96
Opening Statement
The seemingly frantic development of technology for radiation therapy that followed close on the
heels of 3D-CRT and IMRT continues unabated for reasons that are at best speculative. However,
two things are clear: (a) there is no evidence that IGRT, respiratory gating, electronic portal
imaging, ultrasonic guided patient positioning, nor even IMRT itself have led to increased
survival times; (b) the cost of radiation therapy has risen in parallel. In what can best be
described as a blind gallop towards increasingly more precise means of tumor localization,
physicists and manufacturers appear ignorant of the vast experience with cobalt-60, linear
accelerators, and betatrons that began accumulating in the 1950's. Do they actually believe that
failures to achieve local control were due to consistent patterns of geographical miss, or that a
gain in the therapeutic ratio is all it will take to improve outcomes?
If future developments in radiation therapy are to significantly improve outcomes, then these
developments must impact on those cancers that cause the largest numbers of deaths which, in
the USA, are: lung (160 390), digestive system (134 710), breast (40 910), and prostate (27
050).1
From 1977–2002, the five-year relative survival for all stages of lung cancer went from 13% to
16%.1 Surgical resection remains the treatment of choice, while radiation is employed mainly for
the medically inoperable but with much poorer results.2 As for dose escalation, using 3D-CRT
Kong et al.3,4 found that in going from 74–84 Gy to 92–103 Gy, overall survival in a small
number of patients increased by a statistically insignificant amount, 22% to 28%, but with
acceptable levels of toxicity.
Digestive tract cancers are treated mainly by surgical resection with pre- or post-operative
irradiation and chemotherapy. Relative five-year survivals range from 65% for colorectal, 14%
for esophagus, to 5% for pancreatic cancers. 1 When viewed in terms of the overall treatment
strategies for lung and digestive tract cancers, there is nothing to suggest that techniques beyond
CT localization and 3D-CRT will improve survival.
The five-year relative survival for breast cancer increased from 75% to 89% over the period
1977–20021 due mainly to screening and the detection of carcinoma in situ, and small lesions
having a low incidence of positive nodes. Radiation following lumpectomy has reduced local
recurrence and most of the currently employed irradiation techniques are equally efficacious in
reducing cardiac morbidity.
Whether by prostatectomy or 3D-CRT, the current ten-year cause-specific survival for early-
stage prostate cancer is 90% or better.5,6 Late-stage disease does not fare as well but this is hardly
due to poor tumor localization. Since this is a disease of older men, 80–85% of men with prostate
cancer die of other causes.7 There is little reason to believe that further refinements in tumor
localization, more precisely defined dose distributions, or dose escalation, will affect current
levels of outcome.
Despite the myriad technical advances over the past decade, their contributions to survival rates
are undetectable, albeit there have been reduced levels of toxicity in some cases. It is far more
likely that improvements will come from sensitizing drugs and chemoradiation.
AGAINST THE PROPOSITION: Dirk Verellen, Ph.D.

97
Opening Statement
The Proposition is without doubt true and, in fact, need not be limited to future developments
only. A review of the Swedish registry8 reveals that cure rates have not improved over the last 30
years stage-by-stage for most solid tumors. Relative survival rates in NSCLC are identical for
patients diagnosed between 1964–1966 and 1994–1996, regardless of technological
improvements. For breast cancer an improvement has been observed, but this might well be
attributed to the Will Rogers phenomenon in that tumors are discovered sooner and tumor
volume has decreased stage-by-stage, apparently improving outcome.9 Apart from the
introduction of CT-based dose planning, radiation oncology never experienced revolutionary
steps forward in the improvement of treatment outcomes during the past decades. The history of
radiotherapy has followed a slow but distinct path of evolutions, each of which represented some
improvement. We have to acknowledge that there exist almost no randomized trials with
clinically relevant end-points (such as overall survival) proving an evidence-based benefit from
these technical innovations, although complications have been reduced. As cancer becomes a
chronic disease, quality of life should be reconsidered in the ranking of relevant end-points.
Returning to the example of NSCLC, the collaborative group 10 showed that chemotherapy
combined with radiotherapy outperformed radiotherapy alone, but introducing a different
fractionation schedule (CHART) made an even larger improvement in outcome (without altering
the chemotherapy).11 The latter, however, came at the cost of complications. Imagine the results
if these studies would have been performed with the latest IGRT technology: improved outcome
with reduced complication rate.
Developments in radiotherapy are technology driven: energies escalated from kV to MV;

customized blocks replaced standard shielding which, in turn, were replaced by computer-
controlled MLCs; introduction of arc-therapy; 3D-CRT has given way to IMRT; and heavy ions
promise some interesting radiobiological advantages. One can say that radiation oncology
represents a perfect example of evolution theory in that the most adapted technology survived.
The proposition suggests that the current state-of-the-art in 3D-CRT and IMRT has reached its
summit. But is this true? As the precise dose distributions produced by 3D-CRT and IMRT are
less forgiving in terms of treatment uncertainties these techniques will never reach their true
potential without proper image-guidance, and it seems fair to say that the evolution still
continues.
Finally, most technological innovations focussed on improving spatial selectivity, and IMRT
combined with IGRT might indeed reveal the limits that can be realized with external photon
beam delivery. However, some challenges remain: (1) photons might not suffice to combat radio-
resistant tumors and heavy ions might present a possible technological improvement; (2) tumor
delineation currently represents the weak link in the treatment chain. New imaging modalities
will help to avoid inter-observer variation12 and provide increased functional/biological
information about the tumor in order to focus the treatment more efficiently. These developments
will help us to ―paint dose by numbers‖ by acknowledging the heterogeneous nature of tumors,
which has so far been neglected by delivering homogeneous dose distributions. As always, with
each step forward we realize there is an increased number of things we know too little about. We
continue to evolve.
Rebuttal: Robert J. Schulz, Ph.D.
In his opening statement, Dr. Verellen points out that treatment outcomes have improved little
over the past 30 years. He suggests that ―radiotherapy has followed a slow but distinct path of
98
evolutions, each of which represented some improvement‖ but then goes on to acknowledge that
there is scant evidence linking technical improvements with longer survival time. These
observations point to an issue that should be of great concern to physicians and physicists alike:
if technical developments over the past decade have resulted in improved outcomes, then the
extent of these improvements should be documented.
Admittedly, such documentation will be difficult to come by. When increases in survival are
likely to be small, which they most certainly are in the case of technical developments,
randomized-prospective trials are often impractical because they require large numbers of closely
matched patients accrued over many years. On the other hand, retrospective studies, such as four-
field box versus IGRT for the treatment of prostate cancer, compare what was done yesterday
with something that is being done today. Such studies are on shaky ground because patient
management steadily improves, there are periodic changes and inter-hospital variations in how
the disease is staged, expanded screening results in the detection of earlier-stage, more curable
disease, and newer chemotherapy regimens have come into place.
Lacking such documentation, hypothetical arguments that favor the adoption of complex and
costly systems have gained wide acceptance by physicists and physicians alike. This acceptance
requires a leap of faith that is anything but scientific, while taking our attention away from the
critical issue of costs versus benefits. Until the benefits of technical developments are
demonstrated, their higher rates of reimbursement cannot be justified on clinical or moral
grounds, and such reimbursements rejected by insurers as well as providers.
Rebuttal: Dirk Verellen, Ph.D.
Dr. Schulz misses the point in that adoption of new technology is not a blind gallop but rather an
evolution driven by the willingness to improve quality of health care for each patient. Surgeons
are not challenged to initiate randomized trials proving possible benefit of superior scalpels, nor
did we feel compelled to prove the benefit of portal films (it was accepted as common sense and
good QA). Compared to chemotherapy, radiotherapy is modest in cost (5.6% of cancer care
costs) and far more cost-effective.13 The vast experience that Dr. Schulz refers to cannot compare
to results obtained with the precise dose delivery achievable today. And yes, I would argue that
the large amount of failure in local control could be attributed to geographical miss and poor
therapeutic windows due to the need for large treatment margins. It is with the clinical
introduction of IGRT that we start to understand the true concept of margins and organ motion.
Referring to the example of lung cancer, Dr. Schulz compares apples and oranges with surgical
resection of operable, and radiotherapy of inoperable, lung cancer. Recent studies show
comparable results between surgery and radiotherapy for similar populations, with less co-
morbidity for radiotherapy.14 For breast cancer, Dr. Schulz argues that improvements will more
likely come from chemotherapy, yet provides a nice example where developments in chemo- and
radiotherapy need careful synchronization. Herceptine (trastuzumab) is a case in point offering
52% reduced risk of recurrence (five months increased survival), with a factor of four increase in
cardiotoxicity.15,16,17 A safe combination of this drug with radiotherapy requires a highly accurate
treatment delivery that spares cardiac tissue as much as possible. Dr. Schulz's final statement that
improvement will more likely come from sensitizing drugs and chemotherapy is at best true only
if this is combined with an optimal synergy between surgery and state-of-the-art radiotherapy.
REFERENCES
99
1. A. Jemal et al., ―Cancer statistics, 2007,‖ CA a cancer journal for clinicians 57, 43–66
(2007).
2. H. Johnson and J. S. Halle, ―Radiotherapy for stage I, II non-small cell lung cancer,‖ in
Diagnosis and Treatment of Lung Cancer, edited by F. C. Detterbeck, M. P. Rivera, M. A.
Socinski, and J. G. Rosenman (W. B. Saunders Company, Philadelphia, PA, 2001).
3. F-M. Kong et al., ―High-dose radiation improved local tumor control and overall survival
in patients with inoperable/unresectable non-small-cell lung cancer: long-term results of a
radiation dose escalation study,‖ Int. J. Radiat. Oncol., Biol., Phys. 62, 324–333 (2005).
4. F-M. Kong et al., ―Final toxicity results of a radiation-dose escalation study in patients
with non-small-cell lung cancer (NSCLC): Predictors for radiation pneumonitis and fibrosis,‖
5. G. W. Hull et al., ―Cancer control with radical prostatectomy alone in 1000 consecutive
patients,‖ J. Urol. (Baltimore) 167, 528–534 (2002).
6. M. Zelefsky et al., ―Long-term distant metastases-free survival and cause-specific
survival outcomes after high dose conformal radiotherapy for clinically localized prostate
cancer,‖ Int. J. Radiat. Oncol., Biol., Phys. 66, Issue 3 (Supplement), S9 (2006).
7. A. R. Kagan and R. J. Schulz, ―Intensity-modulated radiotherapy for adenocarcinoma of
the prostate: a point of view,‖ Int. J. Radiat. Oncol., Biol., Phys. 62, 454–459 (2005).
8. M. Talback et al., ―Cancer survival in Sweden 1960–1998—developments across four
decades,‖ Acta Oncol. 42, 637–659 (2003).
9. A. R. Feinstein et al., ―The Will Rogers phenomenon: Stage migration and new
diagnostic techniques as a source of misleading statistics for survival in cancer,‖ N. Engl. J. Med.
312, 1604–1608 (1985).
10. NSCLC Collaborative Group, ―Chemotherapy in NSCLC: a meta-analysis using updated
data on individual patients from 52 randomised clinical trials,‖ Br. Med. J. 311, 899–909 (1995).
11. M. I. Saunders et al., ―Programming of radiotherapy in the treatment of non-small-cell
lung cancer—a way to advance care,‖ Lancet Oncol. 2, 401–408 (2001).
12. J. Van de Steene et al., ―Definition of gross tumour volume in lung cancer: inter-observer
variability,‖ Radiother. Oncol. 62, 37–49 (2002).
13. A. Norlund, ―Cost of radiotherapy,‖ Acta Oncol. 42, 411–415 (2003).
14. J. P. van Meerbeeck et al., ―Randomized controlled trial of resection versus radiotherapy
after induction chemotherapy in stage IIIA-N2 non-small-cell lung cancer,‖ J. Natl. Cancer Inst.
99, 415–418 (2007).
15. Editorial, ―Herceptin and early breast cancer: a moment for caution,‖ Lancet 366, 1673
(2005).
16. D. J. Slamon et al., ―The use of chemotherapy plus a monoclonal antibody against HER2
for metastatic breast cancer that overexpressed HER2,‖ N. Engl. J. Med. 344, 783 (2001).
17. E. Tan-Chiu et al., ―Assessment of cardiac dysfunction in a randomized trial comparing
doxorubicin and cyclophosphamide followed by paclitaxel, with or without trastuzumab as
adjuvant therapy in node-positive, human epidermal growth factor receptor 2-overexpressing
breast cancer: NSABP B-31,‖ J. Clin. Oncol. 23, 7811–7819 (2005).
100
2.9. Compared with inverse-planning, forward planning is

preferred for IMRT stereotactic radiosurgery
Fred Hacker and Daniel Low

OVERVIEW
Stereotactic radiosurgery with IMRT is a complex procedure that demands precision treatment
planning. Many physicists believe that this precision is achieved best by inverse planning in
which one works backward from a desired dose distribution to determine the placement of
intensity-modulated treatment fields. Recent work by one of the authors (F.H.) of this month's
Point/Counterpoint contends that inverse planning is unnecessary, and that satisfactory results
can be obtained by a conventional forward-planning approach. This contention is debated in this
Point/Counterpoint.
Arguing for the Proposition is Fred L. Hacker, Ph.D. Dr. Hacker received his Ph.D. in Plasma
Physics from Dartmouth College in 1994. This was followed by a two year Post Doctoral
Fellowship at the Harvard Joint Center for Radiation Therapy, where he worked on conformal
arc techniques for stereotactic radiosurgery/radiotherapy. Following his fellowship he joined the
faculty at the Joint Center. Since 1999 Dr. Hacker has been the chief physicist for the
Stereotactic Radiosurgery/Radiotherapy Program at the combined department of radiation
oncology for the Dana-Farber Cancer Institute, Brigham and Women's Hospital and The
Children's Hospital Boston. He is an instructor in Radiation Oncology at Harvard Medical School
and is certified by the ABR.
Arguing against the Proposition is Daniel Low, Ph.D. Dr. Low earned his Ph.D. in intermediate
energy nuclear physics at the Indiana University Cyclotron Facility in Bloomington, IN. He
began his medical physics career as a postdoctoral fellow in the Department of Radiation
Therapy at M. D. Anderson Cancer Center, conducting research in the computer optimization
and fabrication of electron-beam bolus. In 1991, Dr. Low joined the Mallinckrodt Institute of
Radiology of the Washington University School of Medicine in St. Louis, MO as an instructor,
where he is currently a tenured associate professor of Radiation Oncology. His research interests
are in the application and dosimetry of intensity modulated radiation therapy and in four-
dimensional imaging. He is certified in Radiation Oncology Physics by the American Board of
Medical Physics and the American Board of Radiology.
FOR THE PROPOSITION: Fred L. Hacker, Ph.D.
Opening Statement
Radiosurgery is the quintessential form of conformal radiotherapy, where a single large dose of
radiation is delivered with submillimeter accuracy to a small intracranial lesion. Traditionally,
linac-based radiosurgery has been delivered with circular collimators using noncoplanar arcs.
Several authors have compared this technique to fixed conformal fields and to inverse-planned
IMRT (IP-IMRT).1,2,3,4 Generally these studies show that circular collimation with arcs is optimal
101
for small ellipsoidal lesions. For larger irregularly-shaped lesions, however, substantial
improvements are achieved with fixed conformal fields, and a further incremental improvement
is achieved through addition of IP-IMRT.
Recently we completed a study evaluating the effectiveness of simplified forward-planned IMRT

(FP-IMRT) for intracranial lesions.5 In this technique initial beam approaches and apertures are
set based on conventional beams-eye-view planning. Isodose distributions are assessed, field
apertures are modified, and a small number of subfields are added to a subset of field approaches.
This process is repeated iteratively until the desired isodose distribution is achieved. We
compared this technique to plans generated with IP-IMRT using dynamic MMLC delivery. Plans
were compared for nine lesions treated with fractionated stereotactic radiotherapy (SRT). Lesions
treated with SRT tend to be larger and more irregular than those treated with radiosurgery, and
should present a greater advantage for IP-IMRT. Plans were evaluated for normal tissue
involvement, target coverage, dose homogeneity and organ-at-risk sparing. For all categories the
IP-IMRT and FP-IMRT plans showed comparable performance. There was no difference in
normal tissue involvement at the prescription level. The IP-IMRT plans had a slight (14%)
advantage in average normal tissue involvement at 80% of the prescription level, but at 50% and
20% of prescription the IP and FP-IMRT plans were within 6%. The FP-IMRT plans had a
marginal advantage in tumor coverage and dose homogeneity. There was no difference in
maximum dose to adjacent organs at risk between the FP and IP-IMRT plans.
While there is little dosimetric difference between IP and FP-IMRT, there are substantial
practical advantages to FP-IMRT, most notably a reduction in required quality assurance. Since
the FP-IMRT plans use a few relatively large subfields, standard hand-calculation models can be
used to verify dose. The complex fluence maps derived from IP-IMRT must be verified with
patient-specific phantom measurements, typically requiring a minimum of 2–3 hours of
additional physics QA. This time can be of critical importance in radiosurgery where a rigid head
frame is used. The FP-IMRT plans should also be less prone to delivery errors caused by small
errors in MMLC characterization and calibration. It has been indicated6 that IP-IMRT plans with
small average leaf separations are particularly susceptible to these errors. Of lesser significance is
a reduction in required monitor units, with corresponding reduction in leakage radiation, and a
reduction in wear on the MMLC. Given these practical advantages to FP-IMRT, and the lack of
any compelling dosimetric advantage for either, FP-IMRT is clearly the preferable choice for
radiosurgery.
Rebuttal
Dr. Low is correct in his assessment of the limitations of traditional radiosurgery techniques
using the Gamma Knife or a linac with circular collimation when treating large irregular lesions.
As he points out, several authors have shown the advantages of using conformal arc or conformal
static field techniques for these lesions. Some have also shown further improvement in moving
from conformal open field techniques to inverse planned IMRT. This, however, does not answer
the question posed in this Point/Counterpoint. The question here is does inverse planning for
intensity modulation provide an advantage over state of the art forward planning incorporating
intensity modulation (FP-IMRT) for radiosurgery.
Intuitively it would seem inverse planning would provide an advantage. In practice, however, we
have not found that to be the case for moderately sized intracranial lesions. One reason for this is
that in radiosurgery many noncoplanar fields are typically used, with their approaches spread
through a large solid angle. In this case the dose distribution characteristics are dominated by
102
beam selection, and intensity modulation primarily fine tunes the high dose isodose distribution.
With a large number of field approaches to choose from, and the rapid dose calculation provided
by most radiosurgery dose algorithms, it is relatively easy to design a selection of subfields that
tune the distribution as desired. When inverse planning is used for these cases we have found that
the individual field fluence maps generated are more complicated, while the composite isodose
distribution is not significantly improved.
Although the potential dosimetric advantage of IP-IMRT can be debated, there are clear delivery
advantages for FP-IMRT. Perhaps the most significant is that the few relatively large subfields
used in FP-IMRT will be less sensitive to errors in MLC delivery, compared with the complex
fluence maps determined through IP-IMRT. Radiosurgery is a one-time event, so reliability of
delivery is critical. That is one reason why the use of circular collimators has persisted, despite
the many studies showing the advantage of field shaping. If one is going to change to a more
complicated treatment technique that is potentially less reliable to deliver, there must be a
corresponding dosimetric benefit. At this point IP-IMRT does not provide that benefit. As new
objective functions and leaf sequencing methods are developed, this may change and the debate
can be reopened. But for now FP-IMRT is preferable to IP-IMRT for radiosurgery.
AGAINST THE PROPOSITION: Daniel Low, Ph.D.
Opening Statement
Stereotactic radiosurgery traditionally uses converging arc beams with cylindrical collimators7 to
treat relatively small target volumes. These arc-based treatments were developed prior to the
introduction of inverse treatment planning, and have the advantage that the entrance and exit
doses are distributed over a relatively large volume. However, they suffer from heterogeneous
target doses, and customization of critical structure avoidance is difficult. There is now a growing
consensus that fixed-field treatments can be used as substitutes for arc delivery,8,9 and that
intensity modulated radiation therapy (IMRT) provides advantages over open field
techniques.10,11,12 For example, Kulik et al.12 recently compared fixed-field IMRT using five
fields against a Cobalt-based irradiator (Gamma Knife) and conventional stereotactic irradiation.
They found that for larger or more complex shapes, intensity modulation improves conformity
and provides lower doses to critical structures compared with conventional approaches. For
relatively large target volumes, the use of intensity modulation may also provide improved
treatment efficiency because conformation of larger targets using conventional stereotactic
treatment planning typically requires sequential deliveries to multiple isocenters.
The answer to the question posed in this Point/Counterpoint may therefore be a function of
target-volume size. For large targets, the question becomes: Is inverse treatment planning
necessary for conformal dose delivery? Of course, development of optimal fluence distributions
requires an objective function and input parameters that allow full use of the steep and relatively
shallow dose gradients along the lateral and depth directions, respectively, for external photon
beams. While the optimal combination of these characteristics has not yet been identified,
evidence shows that dose conformality is possible to an extent not previously attainable, 13
indicating that even with suboptimal algorithms, significantly improved dose distributions are
possible. One reason that IMRT provides highly conformal dose distributions is that it sends a
fraction of the fluence through the critical structures. Reviews of multileaf collimator field
sequences of clinical IMRT treatment plans reveal that for most beams, the treatment planning
system irradiates the critical structures with nonzero fluence. It is unrealistic to expect that a
103
treatment planner could determine the best distribution of fluence to deliver through projecting
critical structures.
For small target volumes, fluence modulation becomes less important than selection of beam
orientations. Kulik et al.12 also showed that beam orientation optimization improves doses to
normal structures relative to larger numbers of preselected orientations. The process of
optimizing beam directions is more complex compared with fluence modulation, because the
influence of reorienting beams on the dose distribution is highly nonlinear. The number of
possible beam orientations rapidly becomes untenable, especially when noncoplanar beams are
allowed, as is typical of stereotactic radiosurgery. For example, if the number of beam directions
is constrained to 201 (the number of individual Cobalt sources on the Gamma Knife), then the
number of directional combinations for 1, 2, and 3 beams becomes 201, 40 200, and 8×106. This
example demonstrates the rapid increase in the available treatment planning space with even a
relatively small number of beams. While many of the theoretically available combinations would
be excluded because of trivial clinical issues (e.g., entering through the lens), the number of
combinations still precludes the manual optimization of beam directions.
The use of manual optimization is appealing because of its apparent simplicity and because of the
historic dependence of radiation therapy treatment planning on previously acquired clinical
experience. Limitations inherent in the use of manual techniques to explore all possible delivery
combinations suggest that manual techniques will eventually be replaced by automated
techniques. As the sophistication of both the treatment planner and commercial planning systems
expands, the benefits of inverse planning will be exploited for the patient's benefit, while quality
assurance processes will ensure patient safety.
Rebuttal
Dr. Hacker makes some excellent points in his argument for the use of forward planning for
stereotactic radiosurgery for relatively large target volumes. Only routine quality assurance is
necessary if the FP-IMRT is delivered using relatively large radiation portals for which
independent monitor unit calculations are accurate. When coupled with dose distributions that are
almost as good as IP-IMRT, the practical advantages of FP-IMRT do indeed win out.
However, any current advantages of FP-IMRT may be short lived. As commercial objective
functions improve and delivery constraints are explicitly considered in the inverse planning
process, both the dose distribution quality and the delivery accuracy of IP-IMRT dose
distributions will improve. Quality assurance of IP-IMRT dose delivery, now requiring
significant manpower, will become more convenient, and will ultimately use automated
calculation-based verification methods. The potential for IP-IMRT to automate the treatment
planning process will eventually make it the more efficient process.
More important is the ability of IP-IMRT to explicitly consider and avoid critical structures that
surround the target. In the case of stereotactic radiosurgery, these may be subregions that are
manually defined or located using functional imaging tests. As the technology of functional brain
mapping improves, we may be able to use this information to preferentially deliver normal brain
doses in an attempt to retain specific brain functions. IP-IMRT will be instrumental in the
treatment planning of these patients.
REFERENCES
104
1. S. Y. Woo et al., "A comparison of intensity modulated conformal therapy with a

conventional external beam stereotactic radiosurgery system for the treatment of single and
multiple intracranial lesions," Int. J. Radiat. Oncol., Biol., Phys. 35, 593–597 (1996).
2. R. M. Cardinale et al., "A comparison of three stereotactic radiotherapy techniques: Arcs
vs. noncoplanar fixed fields vs. intensity modulation," Int. J. Radiat. Oncol., Biol., Phys. 42,
431–436 (1998).
3. S. H. Benedict et al., "Intensity-modulated stereotactic radiosurgery using dynamic
micro-multileaf collimation," Int. J. Radiat. Oncol., Biol., Phys. 50, 751–758 (2001).
4. C. Kulik et al., "Conformal radiotherapy optimization with micromultileaf collimators:
comparison with radiosurgery techniques," Int. J. Radiat. Oncol., Biol., Phys. 53, 1038–1050
(2002).
5. F. Hacker, P. Zygmanski, and N. Ramakrishna, "A comparison of forward planned IMRT
to inverse planned IMRT for stereotactic radiotherapy," Med. Phys. 29, 1367 (2002), abstract.
6. P. Zygmanski and J. H. Kung, "Method of identifying dynamic multileaf collimator
irradiation that is highly sensitive to a systematic MLC calibration error," Med. Phys. 28, 2220–
2226 (2001).
7. W. Lutz, K. R. Winston, and N. Maleki, "A system for stereotactic radiosurgery with a
linear accelerator," Int. J. Radiat. Oncol., Biol., Phys. 14, 373–381 (1988).
8. R. W. Laing, R. E. Bentley, A. E. Nahum, A. P. Warrington, and M. Brada, "Stereotactic
radiotherapy of irregular targets: a comparison be-tween static conformal beams and non-
coplanar arcs," Radiother. Oncol. 28, 241–246 (1993).
9. B. G. Clark, J. L. Robar, and A. M. Nichol, "Analysis of treatment parameters for
conformal shaped field stereotactic irradiation: comparison with non-coplanar arcs," Phys. Med.
Biol. 46, 3089–3103 (2001).
10. H. M. Lu, H. M. Kooy, Z. H. Leber, and R. J. Ledoux, "Optimized beam planning for
linear accelerator-based stereotactic radiosurgery," Int. J. Radiat. Oncol., Biol., Phys. 39, 1183–
1189 (1997).
11. S. L. Meeks, J. M. Buatti, F. J. Bova, W. A. Friedman, and W. M. Mendenhall,
"Treatment planning optimization for linear accelerator radiosurgery," Int. J. Radiat. Oncol.,
Biol., Phys. 41, 183–197 (1998).
12. C. Kulik, J. M. Caudrelier, M. Vermandel, B. Castelain, S. Maouche, and J. Rousseau,
"Conformal radiotherapy optimization with micromultileaf collimators: comparison with
radiosurgery techniques," Int. J. Radiat. Oncol., Biol., Phys. 53, 1038–1050 (2002).
13. K. S. Chao, D. A. Low, C. A. Perez, and J. A. Purdy, "Intensity-modulated radiation
therapy in head and neck cancers: The Mallinckrodt experience," Int. J. Cancer 90, 92–103
(2000).
105
2.10. Very high energy electromagnetically-scanned electron beams

are an attractive alternative to photon IMRT
Lech Papiez and Thomas Bortfeld

OVERVIEW
Very high energy electron beams on the order of 150–250 MeV have potential advantages for
tumor irradiation compared with scanned x-ray beams or photon-mediated Intensity Modulated
Radiation Therapy (IMRT). Some physicists believe that VHEE therapy should be developed
because it (1) presents an interesting technical challenge; and (2) promises to offer improvements
in treating certain tumors. Others feel that the improvements, if any, would be marginal at best,
and that the cost of developing VHEE technology for radiation therapy is not defensible. This
controversy is the subject of this month's Point/Counterpoint.
Arguing for the Proposition is Lech Papiez, Ph.D. Dr. Papiez obtained his Ph.D. in physics from
Silesian University, Katowice, Poland. He was a fellow in theoretical physics at the Dublin
Institute for Advanced Study, Ireland, from 1980 to 1983. In 1984 he moved to Canada where he
worked as a researcher in applied mathematics. In 1989 he enrolled in the medical physics
residency program at the London Regional Cancer Center, Ontario, Canada. In 1992 he moved to
the Department of Radiation Oncology, Indiana University, where he is now Associate Professor
and Director of Radiation Physics. Dr. Papiez's research interests focus on transport theory and
optimal control modeling in radiotherapy applications. Dr. Papiez's section was written with the
assistance of Robert Timmerman, M.D., Colleen DesRosiers, M.S., and Vadim Moskvin, Ph.D.,
from the Department of Radiation Oncology, Indiana University School of Medicine,
Indianapolis, IN.
Arguing against the Proposition is Thomas Bortfeld, Ph.D. Dr. Bortfeld trained in medical
physics at the German Cancer Research Center (DKFZ) and in 1990 received his Ph.D. in
physics from the University of Heidelberg for the development of an early inverse IMRT
planning system. He proved the feasibility of MLC-based IMRT at the MD Anderson Cancer
Center in Houston and continued his career at DKFZ, where he and his research team worked on
the clinical implementation and further refinement of IMRT. Since 2001 he has been director of
physics research in radiation oncology at the Massachusetts General Hospital in Boston, where
he is also an associate professor at Harvard Medical School. Dr. Bortfeld wishes to thank his
colleagues in the department of radiation oncology at MGH for helpful comments.
FOR THE PROPOSITION: Lech Papiez, Ph.D.
Opening Statement
Image Guided 4D IMRT (IG-4D-IMRT) is the next, potentially key, advance in radiation
therapy.1,2 Optimal accounting for all aspects of IG-4D-IMRT will require extremely flexible
delivery of radiation dose. In particular, IG-4D-IMRT will require creation of time-dependent
intensity fields (motion picture fields). While existing MLC-based photon IMRT can deliver still
106
pictures of modulated intensity accumulated over the total time of delivery, it is far from an
optimal tool for IG-4D-IMRT delivery.
Electromagnetic scanning of a keV-level pencil beam of electrons is a proven technique for

motion-picture delivery of electroluminescence intensity in television. A similar technology can
be applied for delivery of IMRT with MeV-level beams of charged particles. The existing
application of scanned proton pencil beam therapy3 demonstrates that this approach can be
applied in clinical treatments. The ideal choice of charged particle for scanning radiation delivery
should be light, cost effective to accelerate and reliable with respect to clinical throughput and
maintenance. Electrons meet all of these criteria, provided that they are energetic enough to
deliver dose to any region of the human body.
Studies of very high energy (200–300 MeV) electron beams (VHEEB) performed to date justify
their utilization in radiation oncology.4,5,6,7 VHEEB are superior to photon beams due to their
specific physical properties of dose deposition, as attested to by the following factors: (1) the
ratio of integral dose to the target, compared with the integral dose to healthy tissue/sensitive
organs, is higher for VHEEB than for photon beams,4,5,6,7 and (2) the absence of electronic
disequilibrium at interfaces for VHEEB therapy avoids under- and over-dosage at boundaries of
organs, and leads to a more uniform dose distribution throughout the target.4,5
VHEEB can be scanned at a speed and intensity that allows simultaneous tracking and IMRT
treatment of moving or deformable targets, a feature absolutely necessary for 4D-IMRT. This
feature is facilitated by high dose rates of VHEEB (of the order of Gy/s), making possible nearly
instantaneous (1–2 sec) delivery of the full dose per fraction per field. It also helps assure
independence of dose contributions to each point from individual pencil beams. 4,5 In contrast to
photon MLC IMRT, VHEEB pencil beam independence ensures that dose calculations for
VHEEB IMRT can be accurately determined through linear (additive) operations. Furthermore,
optimal intensity maps for VHEEB treatments are directly deliverable without being
compromised by contamination and scattering characteristics encountered with photon MLC
treatments.
While an individual VHEEB yields a higher surface dose compared with photon irradiation, this
aspect can be overcome by using multiple directions of delivery and sufficiently high energies.4,5
Higher energy electron beams also address the problem of a large penumbra at depth encountered
in lower energy electron therapy. 4,5 Studies show that the clinical relevance of neutron
contamination associated with VHEEB is negligible. 4,8 Altogether, VHEEB therapy is a very
promising approach in the future of therapeutic radiology.
Rebuttal
Indeed, scanning VHEEB units are currently not in production. However, the idea is not out of
reach. VHEEB would be less expensive than proton therapy. Technology already exists for
building compact, table size (1.9 m×0.75 m) 100 MeV electron microtrons. 9 Progress in
superconducting magnets offers the possibility of further decreasing the cost of producing and
controlling VHEEB. Superconducting magnets, including high magnetic field control/scanning
devices, are being used in research, industry and medicine (MRI), proving that much of the
needed know-how has been developed.
Existing comparisons of VHEEB and photon IMRT in homogenous media6,7 indicate that
improvement in dose sparing of sensitive organs for VHEEB is larger than for MLC-based
107
IMRT, each in comparison with conformal 3D therapy. Physical properties of VHEEB make the
method less sensitive to heterogeneities, facilitating construction of dose distributions in non-
homogeneous tissues that modulation of photon beams cannot achieve.4,5
Most importantly, fast-scanning VHEEB allows realistic, efficient solutions to problems of 4D

image guided radiotherapy. Linac-based MLC beam-intensity shaping is an awkward method for
fast delivery of optimized treatment for moving tissues. Improvements in dose shaping offered by
MLC-based IMRT always come at the cost of decreased efficiency, particularly when tissue-
movement compensation is needed (gating). Only fast electromagnetic scanning of charged
particles offers simultaneous improvements in dose shaping and effective delivery of 4D
radiotherapy.
VHEE may be an acronym not only for "very highly expensive electrons." Before the modern
era, the same abbreviation could have been used to label the automobile, i.e., an unrealistic
concept of "very highly expensive emulation" (of horse and carriage). Opinions were voiced then
that such a "VHEE" was not needed because its construction was expensive, its use uncertain and
horses were cheap, abundant and perfect for transportation. And yet
AGAINST THE PROPOSITION: Thomas Bortfeld, Ph.D.
Opening Statement
Very high-energy electron (VHEE) radiotherapy is not a viable alternative to photon intensity-
modulated radiotherapy (IMRT). The reason is simple: VHEE does not exist, whereas photon
IMRT is a widely available treatment option. The more relevant Point/Counterpoint question is:
does VHEE therapy provide a potential advantage and, if so, is it worthwhile to pursue it? Of
course, with high-energy electrons one obtains a sharper beam, because the increased relativistic
mass, E/c2, leads to less side scatter. The price to be paid for this attribute (besides the expensive
high-energy machine) is a relatively flat depth-dose distribution. Unlike existing clinical electron
beams of 10–20 MeV, VHEE electrons do not stop in the patient and do not provide any sparing
of healthy tissues behind the tumor. To the first order, VHEE with 200 MeV electrons is
dosimetrically equivalent to photon therapy at about 15 MV.
There are some differences between VHEE and IMRT. In soft tissues the lateral dose fall-off is
comparable for VHEE and photons,4 but VHEE is less sensitive to tissue heterogeneity.5 It would
be of interest to perform treatment planning comparisons for cases with substantial tissue
heterogeneity in which the sharpness of photon beams is not quite satisfactory, viz., for lung
treatments.10 A few comparative treatment planning studies have been published for VHEE and
IMRT,6 but they used uniform water-equivalent phantoms. VHEE pencil beams exhibit a
concave depth-dose shape, which is somewhat more favorable in terms of integral dose to normal
tissues compared with the slightly convex depth-dose shape of photon beams. On the other hand,
photons provide better skin sparing than do VHEE electrons. There are several additional issues
related to VHEE that have not been addressed sufficiently in the literature, including questions
regarding beam emittance and neutron contamination.
One thing that is clear about VHEE is that a clinical treatment machine for 250 MeV electrons
would be relatively bulky and expensive. Modern 6–10 MeV electron linacs for photon
production are compact devices (their length can be reduced to less than 0.5 m) that can be easily
mounted onto a gantry, even without bending magnets, or onto a robotic arm, which allows for
108
very flexible irradiation geometries. At energies as high as 250 MeV one should consider using
positively charged particles such as protons or heavier ions because of their good lateral
resolution and excellent depth dose characteristics.
It seems to me that VHEE stands primarily for Very Highly Expensive Electrons. The potential
dosimetric advantages of VHEE are quite small. It has always been difficult to prove a clinical
benefit for a new treatment technique, even when the dosimetric advantages are distinctive.11 In
the case of VHEE, it is unlikely that we would see a clinical benefit even if it were a hypothetical
reality. The high expense to develop and build VHEE radiotherapy facilities are therefore simply
not justifiable.
Rebuttal
Image-guided spatio-temporal IMRT is indeed a very promising development in radiation

oncology. Fortunately, it can be realized with technology that is available today or will become
available very soon; it does not rely on the hypothetical development of VHEE therapy. Tumor
tracking is not absolutely necessary to substantially reduce the impact of motion in radiation
therapy; breath holding and respiratory gating are clinically established methods.12 If the
potential efficiency or precision of tumor tracking is desirable, it can be done well with photons
and MLCs. The leaf speed of 2 cm/s with commercial MLCs is sufficient to follow respiratory
motion.1 There are some additional issues in the case of IMRT, but they can be solved by
straight-forward extensions of current technologies.13
I share Dr. Papiez's enthusiasm regarding scanning beam technology. However, the development
of an electromagnetic scanning system for radiation therapy is not as easy as Dr. Papiez suggests.
In the almost 50 year history of radiotherapy with protons and heavier particles, scanning systems
have been used for patient treatments only in very recent years and only at physics-based
institutions.3,14 From a technical and control point of view, scanning VHEE beams would not be
any easier than scanning heavier charged particles. In realistic cases where the tracking data are
imperfect, scanning with a narrow beam leads to an interplay between the scanning beam and the
remaining motion of the tumor relative to the scanning grid,15,16 which is worse than with MLC
delivery. This effect can be reduced, but not completely avoided, by multiple re-scanning.
The potential of scanned VHEE therapy to deliver dose nearly instantaneously (dose rates of
Gy/s) to a treatment volume might be characterized with one word: scary. Possible point dose
rates of several 100 Gy/s would pose an enormous challenge for the dosimetry and the control
system. Further complicating issues specific to scan-beam delivery of VHEE beams include the
significant bremsstrahlung contribution with maximum photon energies of about 200 MeV,
which almost certainly would create significant shielding problems.
In summary, I do not see a role for scanned VHEE beams. We should rather focus our efforts on
the extension of existing radiotherapy modalities with image-guided spatio-temporal concepts.
Electromagnetic scanning systems for IG-4D-IMRT should utilize the physical (and biological)
advantages of heavier charged particles and therefore be IG-4D-IMPT systems.
REFERENCES
1. P. J. Keall, V. R. Kini, S. S. Vedam, and R. Mohan, "Motion adaptive x-ray therapy: a

feasibility study," Phys. Med. Biol. 46, 1–10 (2001).
109
2. L. Papiez, "The leaf sweep algorithm for an immobile and moving target as an optimal
control problem in radiotherapy delivery," Math. Comput. Modell. 37, 735–745 (2003).
3. E. Pedroni, T. Bohringer, A. Coray, E. Egger, M. Grossmann, S. Lin, A. Lomax, G.
Goitein, W. Roser, and B. Schaffner, "Initial experience of using an active beam delivery
technique at PSI," Strahlenther. Onkol. 175, 18–20 (1999).
4. C. DesRosiers, V. Moskvin, A. F. Bielajew, and L. Papiez, "150–250 MeV electron
beams in radiation therapy," Phys. Med. Biol. 45, 1781–1805 (2000).
5. L. Papiez, C. DesRosiers, and V. Moskvin, "Very high energy electrons (50–250 MeV)
and radiation therapy," Technol. Cancer Research & Treatment 1, 105–110 (2002
6. C. Yeboah and G. A. Sandison, "Optimized treatment planning for prostate cancer
comparing IMPT, VHEET and 15 MV IMXT," Phys. Med. Biol. 47, 2247–2261 (2002).
7. C. Yeboah, G. A. Sandison, and V. Moskvin, "Optimization of intensity-modulated very
high energy (50–250 MeV) electron therapy," Phys. Med. Biol. 47, 1285–1301 (2002).
8. V. G. Nedorezov and Y. H. Raniuk, Photo-division of Nuclei above Giant Resonance
(Naukova Dumka, Kiev, 1989).
9. http://www.dfi.aau.dk/isa
10. M. Engelsman, E. M. Damen, P. W. Koken, A. A. van 't Veld, K. M. van Ingen, and B. J.
Mijnheer, "Impact of simple tissue inhomogeneity correction algorithms on conformal
radiotherapy of lung tumors," Radiol. Oncol. 60, 299–309 (2001).
11. R. J. Schulz and A. R. Kagan, "More precisely defined dose distributions are unlikely to
affect cancer mortality," Med. Phys. 30, 276 (2003).
12. J. Wong, Methods to Manage Respiratory Motion in Radiation Treatment in Intensity-
Modulated Radiation Therapy—The State of the Art, edited by J. R. Palta and T. R. Mackie
(Medical Physics Publishing, Madison, 2003), pp. 663–702.
13. T. Neicu, H. Shirato, Y. Seppenwoolde, and S. B. Jiang, "Synchronizedmoving aperture
radiation therapy (SMART): average tumour trajectory for lung patients," Phys. Med. Biol. 48,
587–598 (2003).
14. G. Kraft, "Tumor therapy with ion beams," Nucl. Instrum. Methods Phys. Res. A 454, 1–
10 (2000).
15. H. Phillips, E. Pedroni, H. Blattmann, T. Boehringer, A. Coray, and S. Scheib, "Effects of
respiratory motion on dose uniformity with a charged particle scanning method," Phys. Med.
Biol. 37, 223–234 (1992).
16. T. Bortfeld, K. Jokivarsi, M. Goitein, J. Kung, and S. B. Jiang, "Effects of intra-fraction
motion on IMRT dose delivery: statistical analysis and simulation," Phys. Med. Biol. 47, 2203–
2220 (2002).
110
2.11. Intensity-modulated conformal radiation therapy and 3-

dimensional treatment planning will significantly reduce the need
for therapeutic approaches with particles such as protons
T. Rockwell Mackie and Alfred R. Smith

OVERVIEW
The advantage of radiation therapy with particles such as protons has been recognized for many
years, and a few institutions have made substantial investments in accelerators for proton
therapy. These advantages yield an improvement in the tumor to normal tissue dose relationship
for protons compared with conventional therapeutic approaches employing high-energy x rays.
However, improvements in the conventional approaches are significantly reducing the gap
between dose distributions provided by proton therapy and those achievable with x-ray therapy.
These improvements include intensity-modulated and conformal radiation therapy, sophisticated
treatment planning systems, and eventually, image-guided intensity-modulated conformal
radiation therapy. Whether these improvements reduce the need for further pursuit of particle
(proton) therapy is the subject of this Point/Counterpoint issue.
Arguing for the proposition is T. Rockwell Mackie. Dr. Mackie is a Professor in the University
of Wisconsin Departments of Medical Physics and Human Oncology with membership in the
University of Wisconsin Comprehensive Cancer Center and the Program in Biomedical
Engineering. He is best known for his work on radiotherapy treatment planning including the
convolution/superposition and the Monte Carlo methods of dose computation. Recently he has
concentrated his efforts on helical tomotherapy, which is the integration of radiotherapy
optimization, intensity-modulated radiotherapy, and tomographic treatment verification. His
career has involved co-founding two companies (Geometrics Corporation and TomoTherapy
Corporation), serving on national committees, and being active in the AAPM.
Arguing against the proposition is Alfred R. Smith. Dr. Smith is a Professor at Harvard Medical
School, and a Biophysicist at Massachusetts General Hospital. He supervises clinical physics for
proton therapy research conducted at the Harvard Cyclotron Laboratory and is Associate
Director for the construction of the Northeast Proton Therapy Center at MGH. He has worked in
neutron (M. D. Anderson Cancer Center), negative pi-meson (Los Alamos Scientific
Laboratory), and multi-leaf conformal (Hospital of the University of Pennsylvania) therapy. At
the NCI he initiated studies for the evaluation of treatment planning for heavy particles, x rays,
electrons, brachytherapy, and radio-labeled antibodies. He is Past President of the AAPM and a
Fellow of AAPM and ACMP.
For the proposition: Thomas Rockwell Mackie
Opening Statement
Proton beam therapy, as practiced today, is much better than conventional photon beam
radiotherapy but often inferior to photon intensity-modulated radiation therapy (IMRT). Because
111
the cost of proton facilities is much greater than for photon IMRT, the cost effectiveness of
photon IMRT is currently much superior.
Utilization of the Bragg peak of proton beams is most effective when there are highly sensitive
tissues immediately distal to the target volume. Unfortunately, this aspect cannot always be used
because of uncertainty in the proton range due to uncertainty in tissue density obtained from CT
scans. Tissue density uncertainty has far less effect on photon beam 3D radiotherapy treatment
planning.
Proponents of proton radiotherapy often do not point out that the lateral falloff in dose, the dose
penumbra, at deeper depths from high-energy proton beams can be less steep than low-energy
photon beams produced by linacs. This means that the beam area has to be made larger than a
photon beam to ensure that a deep-seated target volume is not in the penumbra. The sharp
penumbra of photons is what is being exploited in photon IMRT to conformally avoid
neighboring sensitive tissue structures. Multiple proton scattering causes high-gradient hot and
cold spots to develop if the beam travels through highcontrast heterogeneities or steep surface
irregularities which are typical of head and neck radiotherapy.
Multiple proton scattering also makes junctioning fields tricky. While possible in principle, it is
highly unlikely that we will be making use of proton beam radiography for setup or dosimetric
quality assurance in the near future. By contrast, photon IMRT can account for surface
irregularities and tissue inhomogeneities, can reduce the need for elaborate field junctioning, and
will be capable of geometric and dosimetric verification using an exit detector.
The cost-effectiveness of proton radiotherapy for the vast majority of patients must be
questioned. Results of planning comparisons of conventional proton and IMRT photon
radiotherapy indicate that photon IMRT is often the better plan when the dose distribution
requires a complex conformal shape. Any radiotherapy department can now practice photon
IMRT for a tiny fraction (2 to 5%) of the capital cost of a proton facility. And often, the photon
IMRT patients will get a far better treatment! There are a few cases, such as eye tumors, whereby
low-energy proton beams, will be superior. Unfortunately, there would not be enough cases,
except possibly in the largest managed care networks, to keep a proton facility busy treating
those patients that can really benefit from the expense.
Proton IMRT combined with x-ray verification capability may outperform photon IMRT but at
an even greater expense. Before new proton facilities are built it would make sense to wait until
proton radiotherapy is both capable of verifiable IMRT and its acceleration and beam transport
technology is significantly reduced in price.
In summary, photon IMRT reduces the need for proton beam radiotherapy and other more exotic
particle types. This is because photon IMRT is now more cost effective than proton radiotherapy
in the vast majority of cases.
Rebuttal
My esteemed colleague has confirmed my central point. The major difference between
conventional proton beam radiotherapy and photon IMRT (IMXT) is cost. Professor Smith
related a planning comparison for a nasopharynx case, where for equal NTCP, IMXT had a TCP
of 86% and conventional proton radiotherapy had a TCP of 88%. Given uncertainty in biological
112
models, these results are indistinguishable except that the proton treatment would have cost, by
Professor Smith‘s estimates, 50% to 100% more.
The rest of my rebuttal will examine the issue of intensity-modulated proton radiotherapy
(IMPT). IMPT is technically possible (I am a co-inventor, with Joseph Deasy and Paul Deluca,
on an IMPT patent), however, IMPT will be even more expensive because both beam intensity
and beam energy must be controlled simultaneously. Before IMPT is feasible, more accurate
dose computation is required. Before IMPT is practical, the cost and size of proton accelerator
systems must be reduced.
Perhaps the proponents of conventional proton radiotherapy and IMXT should join forces? The
Bragg peak of a conventional proton beam could be exploited to deliver the bulk of the treatment
dose but be augmented with the addition of IMXT. Junction difficulties could be reduced
wherever possible by making the proton dose gradient in the match region low (and linear if
possible). In this way, IMXT could deliver most of its dose to smaller volumes, near critical
structures and at the boundaries of fields, where their placement could be radiographically
verified. This combination could approach the theoretical advantages of IMPT at a
fraction of the investment in time and money.
Until dosimetry issues are resolved and proton accelerators are reduced in price, IMPT is a
questionable investment. What makes the most technical and financial sense is to concentrate
research and development in IMXT and to adapt IMXT for use in conventional proton facilities
as well.
Against the proposition: Alfred R. Smith
Opening Statement
The proposition ‗‗intensity-modulated x-ray (IMXT) dose distributions (rival those for protons,
and therefore) significantly reduce the need for protons,‘‘ is not supported by basic physics,
treatment planning intercomparisons or clinical results. The proposition is unsupported for
standard proton treatments and even less supported for intensity (fluence)-modulated proton
treatments (IMPT).
Clinical proton beams are characterized by depth-dose distributions having a relatively uniform
low-dose entrance region, followed by a region of uniform high dose (the spread-out-Bragg
peak) at the tumor/target, then a steep falloff to zero dose. In contrast, single x-ray beams exhibit
an exponentially decreasing depth dose after a maximum near the surface, a nonuniform dose
across the target, and appreciable dose beyond the target. Therefore, compared to protons, single
x-ray beams deliver higher doses to normal tissues/organs proximal to the tumor, nonuniform
dose across the tumor, and significant dose to tissues beyond the tumor where the proton dose is
zero. The physical advantage of protons for single beams extends to multibeam treatments,
including intensity modulation. Compared to x rays in comparable beam configurations, protons
will always have superior dose distributions. Moreover, proton beams can be aimed directly at
critical structures lying beyond the target in situations where x-ray beams might be disallowed in
IMXT. As a consequence, for equal normal tissue complication probability (NTCP), protons
deliver higher tumor dose, thus increased tumor control probability (TCP). Conversely, for equal
TCP, protons deliver less dose to normal tissues/organs and hence produce lower NTCPs than do
x rays.
113
Treatment planning intercomparisons have been conducted between IMPT and IMXT for
nasopharynx and paranasal sinus carcinomas, Ewing‘s sarcoma, and non-small-cell lung cancer.1
For nasopharynx, when total NTCPs are normalized to 5%, the total TCP was 97% for IMPT
compared to 86% for IMXT assuming D50(tumor) = 67.5 Gy. Our standard, non-IMPT, proton
plan gave a TCP of 88%. The results for the other clinical sites were similar. Generally, IMPT
plans reduce integral dose by about a factor of two compared to IMXT. It is axiomatic that
normal tissues/organs should not be unnecessarily irradiated.
Using standard, non-IMPT, proton boost treatments, clinical studies have demonstrated: superior
TCP =(80% at 3 years), without visual damage, for paranasal tumors; TCP = 97 and 92% at 5
and 10 years, respectively, for skull-base chondrosarcomas; and superior results in a subgroup of
poorly differentiated prostate tumors. For uveal melanomas, using protons alone, the 5-year
actuarial TCP is 96% within the globe (99% in the high dose volume), with 90% of patients
retaining their eye.
If protons and x-rays cost the same, there would be no justification for using x rays for external-
beam cancer treatments. The additional cost factor for proton therapy over that for high-
technology x-ray therapy is now 1.5–2.0. This will decrease with time. Patients/providers are
paying these additional costs to receive the superior clinical benefits offered by protons. Many
more proton therapy facilities are needed.
Rebuttal
Dr. Mackie states that proton therapy, as practiced today, is often inferior to photon intensity-
modulated therapy (IMXT). We have seen no data supporting this claim. IMXT gives a bath of
irradiation to normal tissues/organs that is not present in standard proton plans, therefore, for the
same NTCP, TCP will be higher for standard protons (target coverage is comparable). More
importantly, IMPT plans will always be superior to IMXT plans. This is dictated by the laws of
physics and is unchangeable. In the new proton facility at MGH, IMPT will be delivered with
scanned pencil beams.
Dr. Mackie‘s descriptions of proton beam delivery are misleading. When beams are pointed
directly at critical tissues/organs immediately distal to the target, protons will irradiate only a
small fraction of the down-stream nontarget volume because they stop, whereas photons will
irradiate all structures in their path distal to the target (and also deliver more dose upstream of
the target). Up to medium depths, lateral penumbra is comparable for photons and protons and,
except for small, spherical volumes, the photon penumbra cannot overcome the advantage of the
stopping and sharp distal fall-off of protons. Proton treatment fields are fully compensated for
surface irregularities, tissue inhomogeneities, and shape of distal target volume. We use an x-ray
tube in the treatment nozzle to set-up and verify treatments. Field junctions and patched fields
are routinely used.
Proton treatment capacity is the problem, not patient accrual. We have a 3–4 month waiting list
for treatments; in our new facility we will treat two shifts/day. Dr. Mackie misstated the relative
expense of photons and protons by a factor of 10 if one compares 30-year costs of facilities
having 3–4 treatment rooms. Patients and referring physicians will make treatment decisions
based on superior clinical results rather than relatively small differences in expense.
1
For nasopharynx, IMPT performed by T. Lomax, PSI, Switzerland and IMXT performed by T.
Bortfeld, DKFZ, Germany. For other sites IMPT and IMXT performed by T. Lomax.
114
2.12. Proton therapy is the best radiation treatment modality for

prostate cancer
Michael F. Moyers and Jean Pouliot

OVERVIEW
Proton therapy for the treatment of cancer has been around for about 50 years, albeit mainly for
ocular tumors. Recently, however, there have been several reports of excellent clinical results
obtained with higher energy beams for the treatment of deeper lesions and these, combined with
the commitment of several manufacturers to provide high-energy proton machines, have led to a
surge of interest in proton radiotherapy. It is likely that the number of proton therapy facilities
worldwide will double in the next few years, so it is important to determine which disease sites
might benefit the most by treatment with this new technology. The results published for the
treatment of prostate cancer with protons have been excellent. This has led some to postulate that
proton therapy is better than any other form of radiotherapy for the treatment of prostate cancer.
Others argue, however, that current therapies are quite adequate and, because of the high cost,
proton therapy for prostate cancer is both extravagant and unnecessary. This is the topic of this
month's Point/Counterpoint debate.
Arguing for the Proposition is Michael F. Moyers. Dr. Moyers received an M.S. in Radiation
Biophysics from the University of Kansas, Lawrence, KS, and a Ph.D. in Medical Physics from
the University of Texas Health Science Center, Houston, TX. Between 1990 and 2005 he was
employed at the Loma Linda University Medical Center where he was a Senior Medical
Physicist and Professor of Radiation Medicine, and was awarded several patents regarding proton
therapy equipment. In 2005 he joined Proton Therapy, Inc. as its Technical Director to build,
educate, and manage multiple particle beam facilities around the world. He was certified in
Radiation Oncology Physics by the ABMP in 2001 and is currently a member of the AAPM
Workgroup on Particle Beams.
Arguing against the Proposition is Jean Pouliot. Dr. Pouliot received an M.Sc. in Experimental
Physics and a Ph.D. in Nuclear Physics from Université Laval, Québec, Canada. From 1993 to
1999 he was a Medical Physicist in the Department of Radiation Oncology at Hotel-Dieu de
Québec and Associate Professor at Université Laval, Québec. He is currently Professor,
Department of Radiation Oncology, at the University of California, San Francisco. Dr. Pouliot's
main research interests are the development and clinical integration of dose-guided radiation
therapy with megavoltage cone-beam computed tomography (CT) for patient verification, organ
motion and tumor evaluation studies during cancer irradiation, and inverse planning for dose
distribution optimization and relative biological effectiveness of image-guided high dose-rate and
permanent prostate implant brachytherapy.
FOR THE PROPOSITION: Michael F. Moyers, Ph.D.
Opening Statement
115
The goals of cancer therapy are to maximize the tumor control probability (TCP) and minimize
the normal tissue complication probability (NTCP). The TCP can be increased by escalating the
tumor dose. The NTCP can be reduced by de-escalating the normal tissue dose and using
noninvasive procedures and, because treatments are provided on an outpatient basis, potential
complications due to hospitalization are eliminated.
In the past, intensity modulated radiation therapy (IMRT) with x rays has been used in an attempt
to escalate the tumor dose and reduce the normal tissue dose compared with traditional three
dimensional (3D) megavoltage x-ray beam therapy. Unfortunately, megavoltage beams deposit
dose maximally near the patient's surface and then continue to deposit dose as they traverse
through to the opposite side. There is little that can be done to affect the total energy deposited
within the patient, i.e., the integral dose. The most that can be achieved using IMRT is to move
dose from one sensitive part of the body to a less sensitive part. In contrast, proton beams deposit
dose maximally at the depth of the tumor and deliver no dose distal to the target. If each x-ray
beam in an optimized IMRT plan is replaced by a proton beam which has been range modulated
to give the same depth-dose distribution over the target volume as the x-ray beam, then it is
obvious that much less dose will be delivered outside of the target volume.1 Although this type of
plan does not take full advantage of a proton beam's characteristics, the exercise does
demonstrate that proton therapy (PT) can always deliver a better dose distribution than IMRT.
For typical PT treatments, the integral dose given to nontarget tissues is only a factor of 0.5–0.6
of that delivered by IMRT.
Brachytherapy (BT) is another modality that has been used in an effort to escalate the tumor dose
and reduce the normal tissue dose. BT is an invasive procedure that has associated risks of
compromising blood vessels, nerves, and other critical tissues. BT also has the risks associated
with anesthesia, infection, and edema. None of these risks is encountered during PT. In addition,
not all patients are candidates for the BT procedure because of preexisting medical conditions
and/or gland size. Another disadvantage of BT is that there is a large range of doses and dose
rates delivered to the tumor cells because of the large inverse square gradients around each
source. This makes the biology of BT difficult to understand and optimize. 2,3 In contrast,
delivering a uniform dose throughout the target with PT is accomplished simply by modulating
the depth of the Bragg peak. Yet another disadvantage of BT is the difficulty in treating the
seminal vesicles and/or lymph nodes if involved. Most often, PT can encompass all three targets
within one field.
As of 2006, over 8000 patients have been successfully treated worldwide for prostate cancer
using PT. The demand for PT is increasing rapidly in part because highly motivated and
informed patients find this modality attractive due to its equivalent cure rate and small risk of
morbidity compared with other treatment modalities. The use of intensity modulated proton
therapy, which is just now becoming available, will further reduce the dose to normal tissues and
may allow further escalations of tumor dose.
AGAINST THE PROPOSITION: Jean Pouliot, Ph.D.
Opening Statement
Proton therapy is a good example of the level of technology the fight against cancer may require.
Add that prostate cancer is the most prevalent type of cancer among men, and one may be
tempted to infer that PT is best for prostate cancer. There are, however, two main reasons that
116
preclude reaching this conclusion. First, prostate cancer patients can count on an armada of well-
proven and long-term validated treatment approaches. And second, there are several significant
unanswered questions related to the technical implementation of PT for prostate cancer.
Very good long-term local control of early-stage disease is obtained with permanent prostate
implant (PPI) brachytherapy. For external beam radiation therapy (EBRT), recent randomized
studies have shown that men with localized prostate cancer have a lower risk of failure if they
receive high- rather than conventional-dose conformal radiation.4,5 When the same high dose is
used, published proton results are nearly identical to EBRT.6 Thus, the key point is that men with
prostate cancer, in particular those with advanced local disease, benefit from dose escalation. A
UCSF7 study shows that the response of the tumor to PPI is more rapid than for EBRT, not a
surprise since the BT dose is substantially higher than the EBRT dose. High dose rate (HDR) BT,
as a boost to EBRT, can also provide significant dose escalation with very good sparing of
organs at risk (OARs).8 Furthermore, the finding that treatment failures of EBRT patients were
located in the proximity of known macroscopic disease, 9 suggests that dose escalation to these
regions would be beneficial. This has been performed with EBRT, PPI and HDR, 10 where
increasing the dose to the (MR spectroscopy defined) dominant intraprostatic lesion (DIL) was
achieved without increasing dose to OARs.
Can PT safely boost a DIL within the prostate? Does PT produce the most conformal dose
distribution? The physics can indicate the theoretical potential of each technique: Bragg peak vs
inverse square law (and short distances) vs IMRT. But in the end, it depends on the specific
clinical implementation. For instance, the prostate moves from day to day, sometimes from
minute to minute. Anatomical changes such as rectal filling may have a large impact on local
control when using EBRT.11 Therefore, where the dose is delivered is of equal importance to
dose escalation itself. Prostate motion and setup uncertainties are not an issue for BT, and for
EBRT are accounted for by precise daily alignment of the prostate using markers or other forms
of image guided radiation therapy (IGRT). Does PT resolve the organ motion problem? There are
technical problems of using high-Z markers in proton beams, especially with a small number of
beam angles. Daily rectal balloons can be tolerated by patients during PT, but they may push the
rectum into high dose regions, and residual prostate motions are observed. What are the
consequences of a sudden Bragg peak displacement due to bowel or rectum fillings being
replaced with gas? Because of the sharp dose falloff, it is even more important to integrate the
lessons learned with IGRT in PT before its benefit for prostate may be realized. There are other
questions: economics, neutron dose, biology, etc.
PT may be equivalent in some prostate cases to other radiation therapy modalities, but the answer
today to the Point/Counterpoint Proposition is clearly NO. Perhaps PT might offer a unique
approach for a subset of prostate patients, or even be the best approach in the future. It is only by
performing more research and pursuing more developments that future prostate patients will have
their hope of being cured increased while enjoying a high quality of life.
Rebuttal: Michael F. Moyers, Ph.D.
Dr. Pouliot's first argument against PT is that there are many well-proven and long-term
validated treatment approaches for prostate cancer. While this is true, there is still room for
improvement in TCP and NTPC. IMRT and BT have already been highly optimized and further
technical improvements in these delivery approaches are not obvious. Indeed, Dr. Pouliot's
reference showing the benefits of dose escalation was performed with proton beams.5 The low
rate of complications in this study indicates that even further dose escalation is possible with PT.
117
The basic technique for PT is two uniform lateral beams. 12 If advanced imaging techniques could
demonstrate a localized region of the prostate with a high concentration of tumor cells, then a
boost dose could easily be provided by strategically placing proton Bragg peaks at the site of
those tumor cells using narrow beams from only one or a few directions that bypass critical
normal tissues.
Dr. Pouliot's second argument concerns a number of perceived technical issues, most of which
have been answered over the last 20 years. All prostate patients that have received PT were
localized daily with orthogonal kilovoltage x rays. In fact, IGRT has always been an integral
component of PT. Most patients had the prostate immobilized between a urine-filled bladder and
a constant-volume water-filled balloon inserted into the rectum. This technique, which was well
tolerated, displaced gas cavities from the path of the protons, and pushed most of the rectum and
bladder out of the irradiated volume.13,14 A study that compared a set of four implanted marker
seeds with skeletal landmarks demonstrated that both intra-fraction and inter-fraction motion
were minimal using this immobilization technique. 15
With regard to neutron dose, a recent study has shown that, in the plane of the isocenter, the
neutron dose equivalent from an entire course of treatment is roughly equivalent to that received
from a single CT exam.16
In conclusion, there appear to be no impediments to dose escalation using PT while maintaining a

low risk of complications.
Rebuttal: Jean Pouliot, Ph.D.
I agree with Dr. Moyers that some of the inherent theoretical physics characteristics of proton
beams are of prime significance for prostate cancer treatment. However, in an era of evidence-
based medicine, the best modality provides the best disease control with the least complication
rate at the lowest cost, along with the best global quality of life.
Studies show that what is most invasive for one patient may be least for another. For the same
treatment outcome, completing a treatment in one or two days, even with needle insertion, may
sound appealing compared with several weeks of daily commuting in order to receive a
fractionated course of treatment. Years of improvements on the practical issues have led to
sophisticated treatments using either IMRT or BT, which can address the practical issues for each
specific cohort of patients. PT is in its infancy compared to IMRT or BT, and if you can afford
the price tag, it can offer an equivalent treatment outcome. IMRT and BT are also evolving. In
many ways, we are faced with a moving target. A fair evaluation should compare the best of all
three modalities.
If there is one thing that I remember of my transition from nuclear to medical physics, it is that a
solid fundamental concept is only the beginning of a long uphill road leading to an improvement
in care. Radiation Oncology is a multidisciplinary field with, at its core, a unique human being in
need of treatment. One should devote all efforts to perfect the treatment today to make it better
than the one of a few years ago, while still realizing that new knowledge may eventually make
today's treatment obsolete, and the sooner the better.
REFERENCES
118
1. M. A. Urie, A. Niermerko, A. R. Smith, ―Optimized photons vs. protons,‖ Particles

(Newsletter of the Particle Therapy Co-operative Group), 13, 6 (1994).
2. R. L. Cumberlin and C. N. Coleman, ―New directions in brachytherapy,‖ Int. J. Radiat.
Oncol., Biol., Phys. 53, 6–11 (2002).
3. J. Z. Wang et al., ―Effect of edema, relative biological effectiveness, and dose
heterogeneity on prostate brachytherapy,‖ Med. Phys. 33, 1025–1032 (2006).
4. A. Pollack et al., ―Prostate cancer radiation dose response: Results of the M. D. Anderson
Phase III randomized trial,‖ Int. J. Radiat. Oncol., Biol., Phys. 53(5), 1097–1105 (2002).
5. A. L. Zietman et al., ―Comparison of conventional-dose vs high-dose conformal radiation
therapy in clinically localized adenocarcinoma of the prostate: A randomized controlled trial,‖ J.
Am. Med. Assoc. 294(10), 1233–1239 (2005).
6. M. Roach et al., ―Contemporary results of permanent prostate seed implantation (PPI)
versus conformal proton beam radiotherapy (CPBRT) to 79 Gy,‖ 2005 ASCO Annual Meeting
Proceedings, J. Clin. Oncol. 23, No. 16S (June 1 Supplement), 4639 (2005).
7. B. Pickett et al., ―Three-dimensional conformal external beam radiotherapy (3DCRT)
compared to permanent prostate implant (PPI) in low risk prostate cancer based on endorectal
magnetic resonance spectroscopy imaging and PSA,‖ Int. J. Radiat. Oncol., Biol., Phys. 65(1),
65–72 (2006).
8. I.-C. Hsu et al., ―Normal tissue dosimetric comparison between HDR prostate implant
boost and conformal external beam radiotherapy boost: Potential for dose escalation,‖ Int. J.
Radiat. Oncol., Biol., Phys. 46(4), 851–858 (2000).
9. N. Cellini et al., ―Analysis of intraprostatic failures in patients treated with hormonal
therapy and radiotherapy: Implications for conformal therapy planning,‖ Int. J. Radiat. Oncol.,
Biol., Phys. 53(3), 595–599 (2002).
10. J. Pouliot et al., ―Inverse planning for HDR prostate brachytherapy use to boost dominant
intra-prostatic lesion defined by magnetic-resonance spectroscopy imaging,‖ Int. J. Radiat.
Oncol., Biol., Phys. 59(4), 1196–1207 (2004).
11. R. deCrevoisier et al., ―Increased risk of biochemical and local failure in patients with
distended rectum on the planning CT for prostate cancer radiotherapy,‖ Int. J. Radiat. Oncol.,
Biol., Phys. 62(4), 965–973 (2005).
12. M. F. Moyers and D. W. Miller, ―Range, range modulation, and field radius requirements
for proton therapy of prostate cancer,‖ Technol. Cancer Res. Treat. 2(3), 445–447 (2003).
13. R. B. Ronson et al., ―Patient tolerance of rectal balloons in conformal radiation treatment
of prostate cancer,‖ Int. J. Radiat. Oncol., Biol., Phys. 64(5) 1367–1370 (2006).
14. R. Grover et al., ―Rectal balloons reduce rectal dose in proton radiotherapy for prostate
cancer,‖ presented at the 45th Proton Therapy Co-operative Group Meeting, Houston, Texas,
October 8–11, 2006.
15. R. W. Schulte and T. Li, ―Innovative strategies for image-guided proton treatment of
prostate cancer,‖ Technol. Cancer Res. Treat. 5(2) 91–100 (2006).
16. M. F. Moyers et al., ―Leakage and scatter radiation from a double scattering based proton
beamline,‖ presented at the 45th Proton Therapy Co-operative Group Meeting, Houston, Texas,
October 8–11, 2006.
119
2.13. Proton therapy is too expensive for the minimal potential

improvements in outcome claimed
Robert J. Schulz and Alfred R. Smith

Reproduced from Medical Physics, Vol. 34, No. 4, pp. 1135–1138, April 2007
OVERVIEW
The use of high-energy proton beams for radiotherapy is increasing rapidly, with dozens of new
proton therapy centers planned to open in the next few years. The main advantage of protons over
photons is physical (due to the Bragg peak) since there is little, if any, biological benefit.
However, numerous studies have demonstrated that highly conformal dose distributions can be
obtained with photons using, for example, intensity modulated radiotherapy (IMRT), at
significantly lower cost. One might question, therefore, whether-or-not further improvements in
dose distributions with protons over those already obtainable with IMRT will result in outcomes
that will be improved significantly enough to justify the increased cost. This is the question
Arguing for the Proposition is Robert J. Schulz, Ph.D. Dr. Schulz is a charter member and Fellow
of the AAPM, Fellow of the ACR, and Diplomate of the ABR. His professional career began at
Memorial Sloan-Kettering (1952–1956), developed further at the Albert Einstein College of
Medicine (1956–1970), and concluded at Yale University (1970–1992) from which he retired as
Emeritus Professor. His major contributions have been in radiation dosimetry having chaired the
SCRAD and TG-21 committees, and twice been a recipient of Farrington Daniels Awards. His
retirement to northern Vermont, and close friendship and collaboration with A. Robert Kagan,
M.D., has broadened his perspective on radiation therapy to include considerations of cost versus
benefit, over-diagnosis and over-treatment, the quality of clinical reports, and the impact of new
technologies on outcomes.
Arguing against the Proposition is Alfred R. Smith, Ph.D. Dr. Smith received his Ph.D. in
Physics in 1970 from Texas Tech University. He accepted a postdoctoral fellowship at the M. D.
Anderson Cancer Center in Houston and joined the faculty in 1971. In 1975 he moved to the
University of New Mexico Cancer Center and became Director, Negative Pi Meson (Pion)
Clinical Physics at the Los Alamos National Laboratory. From 1982–1985 he was a cancer
expert at the National Cancer Institute. In 1985 he became Professor, University of Pennsylvania
Medical School, and Director of Clinical Physics, Department of Radiation Oncology at the
Hospital of the University of Pennsylvania. In 1992 he moved to Boston where he held
appointments of Professor at the Harvard Medical School, Biophysicist at the Massachusetts
General Hospital, and Associate Director of the Northeast Proton Therapy Center. He currently
holds appointments of Professor at the University of Texas M. D. Anderson Cancer Center and
Director, Proton Therapy Development. He has served the AAPM as President and Chairman of
the Board of Directors.
FOR THE PROPOSITION: Robert J. Schulz, Ph.D.

120
Opening statement
I submit that the efficacy of any treatment modality depends upon the following: (a) it improves
the rate of cause-specific survival or achieves a similar rate of survival as other treatment
modalities but with fewer complications; (b) it is widely available to patients who will benefit
from it; (c) its benefits outweigh its costs. I further submit that proton therapy (PT), having failed
in each of these categories, should be denied reimbursements beyond those provided for
conventional and equally efficacious treatments.
Although proton dose distributions are unique, except in a few instances the results of PT are
indistinguishable from those achieved by conventional modalities. One exception might be the
treatment of certain large ocular tumors. 1 Whereas small to mid-sized choroidal melanomas are
efficaciously treated by radioactive plaques, 2 for larger tumors (2,400 new cases, 230 deaths in
the USA per year) PT is accepted to be superior. However, ocular tumors require proton-beam
energies no higher than 70 MeV, less than a third of the energy of a general-purpose PT
machine.
It has been widely speculated3,4,5 that PT should be used for the treatment of solid pediatric
tumors (8,400 new cases, uncertain deaths) because of its potential to reduce the impairment of
future development of organs surrounding the tumor. However, there is a paucity of long-term
follow-up data that support these speculations. Also, with advancements in chemoradiation that
require smaller tumor doses, reductions in long-term impairment are equally as likely to be
achieved by 3D-CRT.
As for cancers of the respiratory and digestive systems, and the prostate (680,000 new cases,
330,000 deaths), about the best that can be said for PT is that it is well tolerated.6,7,8 Clearly, any
detectable improvements in the cause-specific survival of these high-profile cancers that could be
attributed to PT would have received banner headlines in the press and medical journals.
There are about 210,000 patients (30% of the 50% of cancer patients who receive radiation
therapy) who, if treated by PT, might do as well as those conventionally treated. Assuming that
three patients are treated per hour in an 8 h day, and each is treated for six weeks, a four-gantry
PT facility could treat 832 patients per year. Therefore, to make PT available to these 210,000
patients would require 232 PT facilities. At a cost of $125 million per facility,9 the capital cost
for these facilities could be $29 billion. As for per treatment costs, these have been estimated 10 to
be 2.4 times higher than IMRT but a factor of three is more realistic when PT is compared with
the actual practice of radiation therapy. Simply put, in the current economic climate the cost-
benefit ratio of PT is unacceptable.
Worldwide, over 47,000 patients have received PT. It would seem that if it had something to
offer patients that x rays don't, something that could justify its high capital and operating costs,
then that something should have long since been recognized.
AGAINST THE PROPOSITION: Alfred R. Smith, Ph.D.
Opening statement
We have insufficient data on costs and benefits for proton therapy to make a definitive statement
about the Proposition. However, I offer some comments about which, I think, reasonable people
can agree.
121
Clinical/biological issues: (1) Dose escalation will increase local control for those tumors that
exhibit dose response;11 (2) Increased local control improves long-term disease-free survival for
some tumors;11 (3) Significant toxic effects in critical normal tissues can limit the potential of
dose escalation, particularly when concurrent chemotherapy is given;12 (4) Highly localized dose
distributions have a greater probability of providing increased local control and decreased late
effects than those with a lesser degree of localization;13 (5) There is no rationale for exposing
critical, uninvolved, normal tissues to radiation when it can be avoided.
Physical properties of treatment beams: (1) Throughout the history of radiation therapy,
improvements in clinical outcomes have been correlated with improvements in dose localization.
Imaging has improved localization of tumors and normal tissues, but this knowledge is of limited
value if one cannot selectively localize the treatment; (2) The ―ideal‖ dose distribution is one that
provides high dose to the treatment target and zero dose to uninvolved normal tissues. Protons
approach the ideal dose distribution to a greater extent than do photons and have about a factor of
2 less integral dose.14
Costs of proton therapy: (1) Capital investment and operating costs for intensity modulated
proton therapy (IMPT) are estimated to be about two times higher than for intensity modulated x
rays;10 (2) Prototype proton therapy systems are expensive but standardization will reduce costs
of individual units (economy of scale) and spare parts inventories. Costs per patient will be
reduced through greater efficiency and fewer treatment fractions. These factors can further
reduce the cost of proton therapy by 30%–40%;10 (3) The high cost of caring for patients who are
not cured, or who suffer injuries from treatment, is usually not considered in cost calculations.
When such costs are included, proton therapy may be less expensive than photons, particularly
for pediatric patients;15 (4) Smaller and more cost effective proton therapy facilities are being
designed.
Proton therapy has significant potential to improve clinical outcomes; this potential has been
borne out in a number of studies.16 It is more expensive than x rays but costs will be reduced over
time. The question is not whether protons will achieve better clinical outcomes, it is ―how much
better‖ and ―is the differential gain worth the additional cost?‖ We have not optimized proton
therapy; IMPT will first be used in the USA in 2007. Multi-institutional clinical trials are being
planned. When we have quantified the clinical benefits of optimized proton treatments, and
established their true costs, we can then solve the cost/benefit equation.
Rebuttal: Robert J. Schulz, Ph.D.
Dr. Smith outlines the precepts for what I'm inclined to call Faith-Based Radiotherapy: if it
seems reasonable, then lack of supporting data should not stand in the way of its implementation.
There is no clearer case of Faith-Based Radiotherapy than PT. Consider that PT began in the
1950s, that there have been as many as 31 clinical programs of which 24 are still active, and that
over 47,000 patients have been so treated (Janet Sisterson, Personal Communication). Yet despite
unimpressive results, its proponents argue that there are insufficient data to arrive at a cost-
benefit ratio, and that unspecified clinical trials are required.
Randomized clinical trials require that the experimental (PT) and control (perhaps 3D-CRT)
groups end up being closely matched. Patient accrual can be slow, and when as with PT the
differences in survival are likely to be small, a rigorous trial would require thousands of patients
122
and take up to ten years if 5-year survival was the end point. With PT being promoted as the
modality du jour, a definitive assessment cannot await the results of a clinical trial.
Protons do to living tissues what x rays have been doing for the past century. Indeed, protons
have unique dose distributions but beyond this their advantage ends, and cost and complexity
weigh heavily against their adoption (just think, port films cannot be made with protons). As for
dose escalation, Dr. Smith's first reference (Kong et al.11) reveals that in going from doses of 74–
84 Gy to 92–103 Gy, the 5-year survival for NSCLC went from 22% (32 patients) to 28% (18
patients). Clearly, this small number of patients distributed over stages I–IIIB cannot be
considered a clinical trial, and provides little support for Dr. Smith's first proposition.
There were 564,000 cancer deaths in the USA in 2006, and nine out of ten were due to metastatic
disease. Are we to believe that PT will have a detectable impact on this statistic?
Rebuttal: Alfred R. Smith, Ph.D.
Dr. Schulz provides (his own) criteria for judging treatment modalities then declares that proton
therapy fails on each point. He sets himself up as judge and jury then arrives at a false verdict. He
asserts:
(a) Proton therapy does not provide the same rate of survival (as x rays) but with fewer
complications. This statement is unfounded.
(b) Proton therapy is not widely available to all who would benefit from it. This is because it is a
new modality, but it is experiencing an impressive growth rate.
(c) The benefits of proton therapy do not outweigh its costs. Unfounded–we have not determined
the overall benefits of proton therapy nor do we know the true costs.
Since the majority of the 47,000 cases that have been treated with protons have been for
choroidal melanomas, Dr. Schulz discredits a potentially important treatment modality for other
cancers before it has been sufficiently tested against standard radiation therapy. Unless the laws
of physics are changed, protons will be superior (in terms of local control, acute and late effects)
to x rays for a broad range of cancer sites, particularly for pediatric cancers. Clinical trials are
being conducted.
He provides unbalanced facility costs and erroneous operating premises. The MGH facility cost
about $42 million. There are smaller proton facilities in final design that will cost less. Facilities
can be used 2 shifts/day, 6 days/week and large ones will treat over 2000 patients/yr. He
misrepresents the conclusions of Goitein and Jermann10 who state: ―The cost of intensity
modulated proton therapy, compared to IMRT, might be reduced to a factor of 1.7 over the next
decade.‖ His ―factor of 3‖ is an invention.
We can understand Dr. Schulz's position on the future of radiation therapy from his quote,
― clinical trials over the past decade suggest that a plateau has been reached and that the
impact of new modalities such as proton beams and IMRT on overall cancer mortality will be
difficult to detect.‖17 He clearly believes that there is nothing we can do to improve clinical
outcomes with radiation therapy—this is in direct contradiction to highly regarded clinicians who
actually treat patients.18,19
REFERENCES
123
1. E. Gragoudas et al., ―Evidence-based estimates of outcome in patients irradiated for

intraocular melanoma,‖ Arch. Ophthalmol. (Chicago) 120, 1665–1671 (2002).
2. S. Packer and M. Rotman, ―Radiotherapy of choroidal melanoma with I-125,‖
Ophthalmology 87, 582–590 (1980).
3. V. C. Wilson, J. McDonough, and Z. Tochner, ―Proton beam irradiation in pediatric
oncology,‖ J. Pediatr. Hematol. Oncol. 27, 444–448 (2005).
4. R. E. Taylor, ―Proton radiotherapy for paediatric tumours: Potential areas for clinical
research,‖ Clin. Oncol. 15, S32–S36 (2003).
5. C. T. Lee et al., ―Treatment planning with protons for pediatric retinoblastoma,
medulloblastoma, and pelvic sarcoma: How do protons compare with other conformal
techniques?‖ Int. J. Radiat. Oncol., Biol., Phys. 63, 362–372 (2005).
6. K. Nihei et al., ―High-dose proton beam therapy for stage I non-small cell lung cancer,‖
7. J. D. Slater et al., ―Conformal proton therapy for prostate carcinoma,‖ Int. J. Radiat.
Oncol., Biol., Phys. 42, 299–304 (1998).
8. T. Hashimoto et al., ―Repeated proton beam therapy for hepatocellular carcinoma,‖ Int. J.
9. D. M. D. Feldstein, ―Anderson private venture raises questions; proton-therapy benefits
at center won't merit costs of care, some say,‖ The Houston Chronicle, Vol. 105. No. 10, Oct. 23
(2005). (http://www.chron.com/CDA/archives/archive.mpl?id=2005_3914112).
10. M. Goitein and M. Jermann, ―The relative costs of proton and x-ray radiation therapy,‖
Clin. Oncol. 15, S37–S50 (2003).
11. F. M. Kong et al., ―High-dose radiation improved local tumor control and overall survival
in patients with inoperable/unresectable non-small-cell lung cancer: Long-term results of a
radiation dose escalation study,‖ Int. J. Radiat. Oncol., Biol., Phys. 63, 324–333 (2005).
12. R. Komaki et al., ―Effects of amifostine on acute toxicity from concurrent chemotherapy
and radiotherapy for inoperable non-small-cell lung cancer: Report of a randomized comparative
trial,‖ Int. J. Radiat. Oncol., Biol., Phys. 58, 1369–1377 (2004).
13. J. Y. Chang et al., ―Significant reduction of normal tissue dose by proton radiotherapy
compared with three-dimensional conformal or intensity-modulated radiation therapy in Stage I
or Stage III non-small-cell lung cancer,‖ Int. J. Radiat. Oncol., Biol., Phys. 65, 1087–1096
(2006).
14. A. J. Lomax et al., ―A treatment planning inter-comparison of protons and intensity-
modulated photon therapy,‖ Radiother. Oncol. 51, 257–271 (1999).
15. J. Lundkvist et al., Cost-Effectiveness of Proton Radiation in the Treatment of Childhood
Medulloblastoma, (2005 American Cancer Society) Published online 6 January 2005 in Wiley
InterScience (www.interscience.wiley.com) 793–801.
16. H. Suit et al., ―Proton beams to replace photon beams in radical dose treatments,‖ Acta
Oncol. 42, 800–808 (2003).
17. R. J. Schulz and A. R. Kagan, ―Weighing proton therapy's clinical readiness and costs,‖
Phys. Today 10–14 (2003).
18. H. Suit, ―The Gray Lecture 2001: Coming technical advances in radiation oncology,‖ Int.
19. B. Glimelius et al., ―Potential gains using high-energy protons for therapy of malignant
tumors,‖ Acta Oncol. 38, 137–145 (1999).
124
2.14. Within the next decade conventional cyclotrons for proton

radiotherapy will become obsolete and replaced by far less expensive
machines using compact laser systems for the acceleration of the
protons
Chang-Ming Charlie Ma and Richard L. Maughan

OVERVIEW
Because they exhibit a Bragg peak, proton beams have been considered potentially superior to
photons and electrons for radiotherapy for over 50 years.1 Yet few proton therapy facilities exist,
primarily because the costs involved are great. Recently, several publications have endorsed the
use of compact laser systems to accelerate protons in place of the very expensive cyclotrons and
synchrotrons that are currently in use. Much research and development is still needed, but some
say that such compact systems may replace conventional accelerators for proton therapy within
the next decade. This conjecture is the topic of this month's Point/Counterpoint debate.
Arguing for the Proposition is Chang-Ming Charlie Ma, Ph.D. Dr. Ma received his Ph.D. in
medical physics from the University of London (UK) and is now Professor and Director of
Radiation Physics at Fox Chase Cancer Center (FCCC). Dr. Ma is active in research, education
and clinical implementation of Monte Carlo simulation techniques for radiotherapy dosimetry,
quality assurance and advanced treatment planning for intensity-modulated radiotherapy and
electron therapy. Since 1999 Dr. Ma has been investigating proton and light ion acceleration
using laser-induced plasmas. He is currently the project leader for laser-proton acceleration in
radiation therapy at FCCC.
Arguing against the Proposition is Richard L. Maughan, Ph.D. Dr. Maughan received his Ph.D.
in physics from the University of Birmingham in England. He started his career at the Gray
Laboratory, London in 1974, and moved to Wayne State University (WSU) in 1983 where he
was responsible for the medical physics aspects of a neutron therapy program. He also was an
active participant in the WSU Medical Physics Graduate Program. He is now Professor, Vice
Chair and Director of Medical Physics in the Department of Radiation Oncology at the
University of Pennsylvania. His research interests are particle therapy (neutrons, protons, heavy
ions), with a particular emphasis on proton therapy.
FOR THE PROPOSITION: Chang-Ming Charlie Ma, Ph.D.
Opening Statement
For treatment of cancer with radiation, the use of proton or light-ion beams provides the
possibility of better target dose conformity and normal tissue sparing, compared with commonly
used photon and electron beams.1 Proton, helium and heavier ion beams were used for
biomedical studies in the early 1950s. The first human patient was treated for a pituitary tumor in
1954, about forty thousand patients have been treated since with proton beams worldwide. Light
125
ion and other particle beams have also shown encouraging results, particularly for well-localized
radioresistant lesions.
Given this advantage, why is proton or ion beam therapy only offered at a few facilities
worldwide? The answer is high expense. Conventional proton or ion facilities are either
cyclotron- or synchrotron-based. An accelerator that is big enough to accelerate protons to the
required therapeutic energies can cost in excess of $50 million. Also, heavy charged particles are
difficult to handle and require a massive gantry for beam delivery. The amount of concrete and
steel used to shield a proton treatment room is an order of magnitude or so greater than for a
conventional linac room. Although use of an expensive accelerator can be made more efficient by
sharing the therapy beams among multiple treatment rooms, this approach requires additional
space for a switchyard housing vacuum beamlines, dipole bending magnets, steering magnets,
focusing quadrapole magnets, power supplies, and cooling equipment. The cost of the gantries
and the building increases the total capital cost to about $100 million for a proton facility, and 2–
3 times that for an ion facility.
Laser acceleration was first suggested for electrons. 2 Rapid progress in laser-electron
acceleration followed the development of chirped pulse amplification and high fluence solid-state
laser materials such as Ti:sapphire. Recently, charged particle acceleration using laser-induced
plasmas has illuminated the search for a compact, cost-effective proton or ion source. Proton
acceleration is achieved by focusing a high-power laser pulse onto a thin target. The short
(femtosecond) pulse width of the laser produces a high peak intensity that causes massive
ionization in the target and expels a great number of relativistic electrons. This sudden loss of
electrons leaves the target with a highly positive charge, yielding an effective electric field of
~1012 V cm–1. This transient field then accelerates protons and other light-positive ions, if
present, to high energies. Laser-proton acceleration has been investigated by major laboratories
worldwide, and energetic protons up to 58 MeV have been generated using high-intensity, short-
pulse lasers.3 Theoretical studies show that at a laser intensity of 1021–1022 W cm–2, protons
may be accelerated up to 300 MeV with a spectrum and angular distribution. 4 An experimental
facility dedicated to laser-proton acceleration for cancer treatment has recently been established5
with support from the U.S. Department of Health and Human Services. Because of the small
acceleration distance, a laser-proton/ion accelerator will be much more compact than
conventional cyclotrons or synchrotrons. Once developed, the laser-proton ion accelerator may
be the best candidate for particle therapy.
AGAINST THE PROPOSITION: Richard L. Maughan, Ph.D.
Opening Statement
At the present time laser accelerators for use as electron, proton or heavy ion accelerators are in
an early stage of research and development.
For proton acceleration, with potential use as a source for proton therapy, the best published data
are those presented by Malka et al.6 Using a 10 Hz laser system with a peak irradiance of
6×1019 W cm–2, they have accelerated protons in a broad spectrum from 0–10 MeV. They
estimate that a peak irradiance of 2×1021 W cm–2 is required to accelerate protons to the energies
required for proton therapy, (i.e., a laser at least 30 times more intense that used in their
experiment). The highest reported proton energies3 are 58 MeV. The ability of laser accelerators
to produce sufficient dose rate has not yet been demonstrated. Lasers producing beams of this
peak intensity have yielded only a single laser pulse approximately every 20 minutes There are
126
no reports of laser proton beams being transported from the proton producing target along a beam
line to a remote experimental location. That is, laser produced proton beams have not yet been
used in any practical application.
Laser produced proton beams have a broad energy spectrum which is problematical.4 A magnetic
spectrometer has been proposed to produce a pseudo monoenergetic beam.7 Even after removal
of unwanted protons (~99.5%) in this manner, the energy spread on a laser-produced proton
beam is still inferior to a conventionally-accelerated proton beam, leading to a significant
degradation in the sharpness of the beam's distal edge. The 99.5% of the beam that is stopped in
the spectrometer will present considerable shielding challenges, due to neutron production and
materials activation. A detailed analysis of this problem has not been published, but the shielding
requirements may be a significant impediment to mounting the laser on a gantry.
Laser pulses are only a few ps long, so there is no opportunity for beam scanning during the pulse
and, at best, a laser may produce only up to 600 pulses/minute. Even if the intensity and energy
of each pulse could be accurately controlled (problems which have not yet been addressed),
600 pulses/minute are insufficient for delivery of a spot-scanned treatment in a reasonable time
(1–2 minutes). Only scattered beam delivery would be possible, and not only would this
approach yield an inferior distal edge, but also it would not be capable of conforming to the
proximal tumor edge. A scattered system requires a long beam path (~2 m for divergence and to
improve the Bragg peak-to-plateau ratio; this requirement, combined with the magnetic
spectrometer and the necessary shielding, will produce a gantry of similar proportions to a
conventional gantry. Thus, there will be little or no cost savings on the gantries, which constitute
a major part of the cost of a multi-room proton therapy facility. Hence, the cost and space
requirement advantages of a laser accelerator are questionable.
For all these reasons it is unlikely that laser beam technology will replace conventional
acceleration techniques in proton therapy facilities within the next decade.
Rebuttal: Chang-Ming Charlie Ma, Ph.D.
Radiation therapy has experienced revolutionary changes over the last 100 years, ranging from
kilovoltage x-ray therapy to isotope teletherapy, megavoltage x-ray and electron therapy. From
time to time, technological breakthroughs have brought marked improvement in treatment
efficacy. Laser-proton acceleration is still in its infancy. Like many other new findings, its
success depends on the advancement of science and technology. Significant progress has been
made in laser physics in the last decade including new laser materials, novel amplification
methods and compact, high-power laser systems. A number of petawatt lasers are being built
worldwide with higher rep rates and intensities (1021 W cm–2 or higher). Based on particle in cell
simulations, the dose rate of protons using such lasers may reach 10+ or even 100+Gy/min.7
Because of the broad energy spectrum of laser-accelerated protons, it is easier to obtain a desired
spread-out Bragg peak by selecting protons across a narrow range of energies. The shielding for a
laser-proton treatment gantry may not be much more than that for a conventional proton gantry,
because most unselected protons have energies well below 50 MeV. The broad angular
distribution of laser-accelerated protons will actually provide superior dose profiles without the
need for large SSDs. Beam scanning is an effective way for proton beam intensity modulation
but it is not the only way. For example, conformal proton dose distributions can also be achieved
using multileaf collimators and port-specific compensators via aperture-based plan optimization.8
This method requires a lower dose rate than the beam-scanning method. A laser can also serve
multiple treatment rooms with little additional cost, since the transport of laser light is much
127
easier to achieve than is the transport of protons. It is expected that many scientific and
engineering issues must be resolved before laser-accelerated protons can be used for cancer
therapy, but they also represent exciting challenges for future research.
Rebuttal: Richard L. Maughan, Ph.D.
With present technology the capital costs of proton therapy are certainly high, but it is not clear
that laser accelerators can significantly reduce overall facility costs.
Regardless of the proton acceleration method used, most components of the facility remain
unchanged. The cost of a proton gantry with bending magnets, a beam nozzle and patient support
system is high. Scattered proton beams require a gantry of large dimensions, because double-
scattering systems require a path length of 2.5 to 3 meters to spread the beam to clinically useful
dimensions. These space requirements are the same for all accelerators. The laser system requires
an energy selector magnet with neutron shielding and, therefore, gantries may be similar in mass
and cost to existing gantries. A laser accelerator needs to produce a proton beam of the same
energy and intensity as existing beams, so with similar gantries, treatment room size and
shielding requirements will be identical. Hence, at best, the only cost savings are on the beam
switchyard and the accelerator, which are a small proportion of the total facility cost. Although a
research laser may be of modest price ($1M), a fully-developed, FDA-approved, commercial
laser accelerator, with all the necessary safety features, will probably cost 3 to 5 times more.
Thus, a laser-based facility could cost $85M or more, in comparison with an existing facility
costing $100M.
Recent laser experiments on plasma channeling demonstrate the possibility of producing quasi-
monoenergetic particle beams,9 thus offering a potential solution to the energy selection and
neutron shielding problem. This may allow for development of a lightweight, less expensive
gantry-mounted laser accelerator for installation in a single shielded room. This would be a
significant advance for proton therapy, but one that is unlikely to occur within the next decade.
REFERENCES
1. R. R. Wilson, "Radiological use of fast protons," Radiology 47, 487–491 (1946).

2. T. Tajima and J. M. Dawson, "Laser electron accelerator," Phys. Rev. Lett. 43, 267–270
(1979).
3. R. A. Snavely, M. H. Key, S. P. Hatchett, T. E. Cowan, M. Roth, T. W. Phillips, M. A.
Stoyer, E. A. Henry, T. C. Sangster, M. S. Singh, S. C. Wilks, A. MacKinnon, A. Offenberger,
D. M. Pennington, K. Yasuike, A. B. Langdon, B. F. Lasinski, J. Johnson, M. D. Perry, and E.
M. Campbell, "Intense high energy proton beams from petawatt-laser irradiation of solids," Phys.
Rev. Lett. 85, 2945–2948 (2000).
4. E. Fourkal, B. Shahine, M. Ding, J.-S. Li, T. Tajima, and C.-M. Ma, "Particle in cell
simulation of laser-accelerated proton beams for radiation therapy," Med. Phys. 29, 2788–2798
(2002).
5. C.-M. Ma, E. Fourkal, J.-S. Li, B. Shahine, M. C. Lee, S. McNeeley, W. Xiong, W. Luo,
and A. Pollack, "A Laser-Proton Therapy System for Radiation Oncology," Proc. of the XIVth
International Conference on the Use of Computer in Radiation Therapy (ICCR, Seoul, 2004),
pp. 177–182.
128
6. V. Malka, S. Fritzler, E. Lefebvre, E. D'Humières, R. Ferrand, G. Grillon, C. Albaret, S.

Meyroneinc, J.-P. Chambaret, A. Antonetti, and D. Hulin, "Practicality of protontherapy using
compact laser systems," Med. Phys. 31, 1587–1592 (2004).
7. E. Fourkal, J. S. Li, M. Ding, T. Tajima, and C.-M. Ma, "Particle selection for laser-
accelerated proton therapy feasibility study," Med. Phys. 30, 1660–1670 (2003).
8. C.-M. Ma, I. Veltchev, E. Fourkal, J. S. Li, W. Luo, J. Fan, T. Lin, and A. Pollack,
"Development of a Laser Driven Proton Accelerator for Cancer Therapy," Laser Physics (in
press).
9. S. P. D. Mangles, C. D. Murphy, Z. Najmudin, A. G. R. Thomas, J. L. Collier, A. E.
Dangor, E. J. Divall, P. S. Foster, J. G. Gallacher, C. J. Hooker, D. A. Jaroszynski, A. J. Langley,
W. B. Mori, P. A. Norreys, F. S. Tsung, R. Viskup, B. R. Walton, and K. Krushelnick,
"Monoenergetic beams of relativistic electrons from intense laser-plasma interactions," Nature
(London) 431, 535–538 (2004).
129
2.15. High energy electron beams shaped with applied magnetic

fields could provide a competitive and cost-effective alternative to
proton and heavy-ion radiotherapy
Frederick D. Becchetti and Janet M. Sisterson
OVERVIEW
Work at several institutions has demonstrated that intense magnetic fields can be used to confine
and shape high-energy electron beam-dose profiles. It is conceivable that this approach might
compete effectively with hadron therapy. However, the technique gives rise to several technical
challenges, some of which may be insurmountable. Whether the technique has enough potential
to be pursued as a possible approach for radiation therapy is the subject of this month's
point/counterpoint debate.
Arguing for the Proposition is Frederick D. Becchetti, Ph.D. Dr. Becchetti has worked primarily
in accelerator-based nuclear physics including the application of large super-conducting magnets
to nuclear reaction studies and recently, application of the latter to radiation oncology. He has
served on the Board of Editors of Review of Scientific Instruments (1999–2001) and was
associate program chair for the recent Symposium on Radiation Measurements and Applications
held in Ann Arbor, Michigan, May 21–23, 2002. Dr. Becchetti has chaired or co-chaired a
number of Ph.D. thesis committees dealing with medical imaging, radiation oncology or related
areas in nuclear medicine.
Arguing against the Proposition is Janet M. Sisterson, Ph.D. Dr. Sisterson trained as a Medical
Physicist in London following receipt of a Ph.D. in high-energy physics from the University of
London. She spent 25 years at the Harvard Cyclotron Laboratory where she helped develop many
of the techniques used worldwide in proton radiation therapy. In 1998 she moved to the Northeast
Proton Therapy Center, Department of Radiation Oncology at the Massachusetts General
Hospital. She holds a joint appointment as Assistant Professor in the Department of Radiation
Oncology at Harvard Medical School. She is the principal investigator on a NASA grant to
measure proton and neutron cross sections needed for cosmic-ray studies.
FOR THE PROPOSITION: Frederick D. Becchetti, Ph.D.
Opening Statement
Calculations1 and measurements using phantoms2,3 have shown that the radiation dose profile
from high-energy electron beams typically used in radiotherapy can be confined using high (viz.,
a few tesla) magnetic fields. The large penumbra associated with scattering of electron beams in
tissue-like material is thereby greatly reduced. In addition, for magnetic fields parallel to the
initial beam direction, additional, 3D focusing of the electron beam may be obtained. The net
effect is a greatly enhanced dose near the end of the electrons' range. The dose profile resembles
a degraded Bragg curve.2,3 This feature could be exploited in conformal radiotherapy to furnish
an alternative to proton or heavy-ion radiotherapy, at least for certain types of tumors, and to
130
provide an economical, on-site electron-beam therapy facility for many hospitals. High-energy
electron accelerators (e.g., linacs and microtrons) and electron-beam gantries are affordable by
many hospitals. This is in contrast to current proton and heavy-ion radiotherapy units, which are
large, costly facilities located far from most hospitals and hence not practical for widespread use,
especially when fractionated doses are needed. Many of the technical problems have been solved,
particularly the use and efficient operation of large-bore superconducting magnets and such
magnets are in routine use in MRI units at most large hospitals. Such magnets can be designed, if
needed, to allow rotation with a treatment gantry.4 Recently mechanical coolers have been
utilized which can further simplify a gantry-mounted system, and nonmagnetic treatment gantries
and tables are feasible. Further research and development appears warranted and if successful
could lead to animal and, eventually, human clinical trials in the near future.
Rebuttal
It is not necessarily claimed that electrons confined by a magnetic field are preferable to high-
energy protons (or heavy ions) for many types of radiotherapy. Instead, as noted, the high cost of
high-energy ion accelerators, gantries, radiation shielding and buildings has greatly limited the
number of patients that can be treated with protons. For some patients, magnetically-confined
electron-beam therapy may in the future provide a viable alternative to protons at lower cost. The
experiments cited in Refs. 2 and 3 utilized a hospital-based "table top" 50 MV electron racetrack
microtron. 100 MV electron microtrons used for physics research are similarly in operation.
Providing an accurate 3D focus for stereotactic treatment with electrons is perhaps the major
challenge. As noted in the opening statement, large-bore superconducting magnets (e.g., 1 m
diam. bore×1 m long B 5T) are feasible, and can be gantry mounted. Such magnets are
presently used in research and industrial applications as well as in medical applications (high-
field MRI). Thus much of the needed engineering has already been done. The key issue is to
demonstrate that magnetically-confined electron-beam radiotherapy may have advantages over
other modalities, even x-ray IMRT in certain circumstances. 1,2,3,5 However, this issue can be
addressed definitively only by further research including realistic treatment simulations and
animal clinical trials.
AGAINST THE PROPOSITION: Janet M. Sisterson, Ph.D.
Opening Statement
Computer simulations and some experiments show that intense magnetic fields can be used to
delineate the geometry and control the dose distribution of high-energy electron beams. For this
reason, electron beams have been proposed as a cost-effective alternative to heavy ion beams,
with a similar therapeutic benefit. Simulations and experiments have been conducted in an
attempt to overcome the inherent limitations of electrons, all due to the small electron mass, and
to mimic, by external means, the inherent properties of ion beams. The goal of these experiments
has been to identify applications where electron beams can be equally or preferentially used to
treat specific tumor sites.
The inherent physical properties of ion beams are used to advantage in radiation therapy. These
properties include: (1) little penumbral scattering as the beam penetrates tissue; (2) a finite range
in tissue, accurately controlled by the beam energy; (3) a very sharp fall-off in the distal edge of
the dose distribution; and (4) a maximum rate of energy loss near the end of range: "the Bragg
peak effect." These properties are used to achieve dose distributions that are uniform over large
131
target volumes, with a sharp penumbra in all directions perpendicular to the beam and along the
axis. The resulting dose distributions conform closely to the target volumes, leading to maximum
sparing of adjacent normal tissues and sensitive structures.
Electrons, as continuously ionizing particles, share some of the properties of heavy ions. In
electron beams, however, the benefit of these properties is lost because of the extreme scattering
of individual electrons. Intense magnetic fields can limit the scattering effects to a significant
degree. Still, they do not allow electrons to achieve the properties of an ion beam Bragg peak or
the dose distributions attainable by summation of such peaks. For example, the proposed
magnetic fields focus primarily in one dimension. This implies that any magnetic focusing
technique will require much more sophisticated, three-dimensional control of the magnetic focus
inside the patient. In addition, electron energies up to 100 MeV may be needed to cover the full
range of therapeutic applications. The construction of a device that delivers high-energy electron
beams with a precisely controlled three-dimensional magnetic focusing lens may prove to be an
insurmountable, and certainly not cost-effective, engineering task.
In contrast, intensity and energy modulated ion therapy is already a reality, and permits precise
control of dose deposition in volumes as large as 20 000 cm3 with a resolution of 0.125 cm3. The
technology for delivering these fields is proven and uses off-the-shelf magnetic devices and
control techniques. Field delivery is accurately controlled in both time and space, a critical
benefit when considering patient-specific concerns such as organ motion.
In summary, one-dimensional magnetic control of electron fields may improve the therapeutic
gain compared with "conventional" electron fields. Such control is no match for intensity
modulated x-ray therapy, conventional ion therapy, or intensity modulated ion therapy. There is
simply not a clear rationale to justify pursuit of such a nontrivial and expensive engineering task
as magnetic field-controlled electron beam therapy.
Rebuttal
Our colleague claims that electron beams, modified with (strong) magnetic fields, can be
delivered and controlled in a clinical setting to produce therapeutic dose distributions that might
be an alternative to proton or heavy ion radiotherapy. Such an alternative is deemed desirable due
to the high cost of a proton or ion facility, and the implied geographic separation of such a
facility from a hospital. We believe there are several misconceptions in this position. First, the
Harvard Cyclotron Laboratory/Massachusetts General Hospital collaboration (HCL), with over
9,000 patients treated, shows that fractionated proton radiotherapy can be successfully
accomplished using a facility located about three miles from the hospital campus. Second, it is
true that proton and ion therapy were pioneered at existing accelerator facilities, often under
many constraints. However, there are currently many hospital on-site proton centers treating
patients, under construction, or planned. Such facilities are expensive but can still be cost-
effective. For example, the cost of such a facility is amortized over the 30-year lifetime of a
cyclotron. (Of note, the HCL operated for 50 years!) Over that lifetime, a conventional clinic,
albeit a large one, would purchase about 12 linear accelerators comparable to the multiple
treatment rooms in a proton therapy facility. This analysis certainly reduces the gap in spending
between a conventional clinic and a clinic that also provides proton therapy. Finally, we again
stress that the predicted electron dose distributions are not comparable to conventional, energy-
modulated, proton Bragg peaks, and certainly not to intensity and energy modulated proton
beams or photon IMRT.
132
REFERENCES
1. A. F. Bielajew, "The effect of strong longitudinal magnetic fields on dose deposition

from electron and photon beams," Med. Phys. 20, 1171–1179 (1993).
2. D. W. Litzenberg, A. F. Benedick, D. L. McShan, T. W. O'Donnell, D. A. Roberts, F. D.
Becchetti, A. F. Bielajew, and J. M. Moran, "An apparatus for applying strong longitudinal
magnetic fields to clinical photon and electron beams," Phys. Med. Biol. 46, N105–N115 (2001).
3. F. D. Becchetti, D. W. Litzenberg, J. M. Moran, T. W. O'Donnell, D. A. Roberts, B. A.
Fraass, D. L. McShan, and A. F. Bielajew, "Magnetic confinement of radiotherapy beam-dose
profiles," Proceedings of Cyclotrons and Their Applications 2001, Sixteenth International
Conference, edited by F. Marti (AIP, Melville, 2001), pp. 44–46.
4. H. Blosser et al., "Superconducting cyclotron for medical applications," IEEE Trans.
Magn. 25, 1746–1754 (1989).
5. J. Chu, W. Hsi, and L. Reiffel, "Potential application of transverse magnetic field in x-ray
therapy," Med. Phys. 28, 1277 (abstract) (2001).
133
2.16. Modern linac stereotactic radiosurgery systems have rendered

the Gamma Knife obsolete
Frank J. Bova and Steven J. Goetsch

OVERVIEW
Precise localization of high doses of radiation in small volumes of tissue has been one of the
strong marketing advantages of Gamma-Knife technology. New linear accelerators, with 0.5 mm
isocentric accuracy and software support for conformal targeting of dose, approach the precision
of Gamma Knife systems. Some physicists think that this precision, combined with the greater
versatility of linacs, make the Gamma Knife an obsolete technology. Others believe that the
exactness of dose delivery with the Gamma Knife ensures that the technology will continue to be
useful. These opposing points of view are explored in this Point/Counterpoint.
Arguing for the Proposition is Frank J. Bova, Ph.D. Dr. Bova attended Renesslear Polytechnic
Institute and in 1972 graduated with a Bachelor in Biomedical Engineering and in 1973 with a
Masters in Biomedical Engineering. He went on to receive a Ph.D. in Nuclear Engineering
Sciences from the University of Florida in 1977. In 1978 Dr. Bova joined the Department of
Radiation Oncology at the University of Florida. In 1991 he was appointed as the Einstein Fund
Professor of Computer-Assisted Stereotactic Neurosurgery within the Department of
Neurosurgery. In January of 1999 he joined the faculty of Neurosurgery at the University of
Florida and was appointed Professor of Neurosurgery.
Arguing against the Proposition is Steve Goetsch, Ph.D. Dr. Goetsch is a self-employed
consultant in radiological physics. He received a Master's degree in Health Physics from
Northwestern University in 1974 and worked as Radiation Safety Officer for Amersham/Searle.
He later returned to graduate school, receiving a Doctorate in Medical Physics from the
University of Wisconsin in 1983. After 7 years at the ADCL, he accepted a position at the UCLA
Medical Center in Radiation Oncology, working there from 1990 to 1994. Since 1994 he has
been Director of Medical Physics at the San Diego Gamma Knife Center. He continues to consult
and do research in stereotactic radiosurgery.
FOR THE PROPOSITION: Frank Bova, Ph.D.
Opening Statement
Any radiosurgery procedure involves three steps: Imaging, planning, and treatment. A
radiosurgery imaging system must include stereotactic imaging and localization systems that
provide sub-millimeter accuracy in specifying the position of target and nontarget tissues. The
treatment planning system must allow the user to develop a highly conformal plan that has rapid
dose falloff for all nontarget tissues, not just the tissues in a specific plane. This is an area where
the plan designed using a linac system can differ from a plan designed using the Leksell Gamma
Knife system. When designing a Leksell Gamma Knife plan, the user encounters a practical
134
limitation in his or her ability to restrict the angles at which the beams can be directed to the
target tissues. Blocking too many beams leads to an unacceptably low dose rate. For linacs, the
user is free to limit the number of beams without affecting the dose rate. If, however, the dose
from noncoplanar beams is spread over less than 400 to 500 arcing degrees, the dose
concentration and falloff are adversely affected. We believe that, regardless of the system, plans
that use too few isotropic beams to deliver the dose to each isocenter violate one of the basic
principles of radiosurgery, namely that all target tissues should experience the full isotropic set of
radiation beams. If the radiosurgical planning system supports isotropic arc placement to target
tissues, then the only questions left for the planning process are a) how large should each
isocenter be, b) how close should the isocenters be spaced, and c) what are the relative intensities
of the various isocenters? These questions are identical whether planning a linac-based
radiosurgery treatment or a Leksell Gamma Knife treatment. Because the questions are identical
irrespective of the radiosurgery system used, an expert user or an automated treatment planning
system can generate identical treatment plans for either system.
The final question is one of plan delivery. Can the treatment system deliver the planned
distribution accurately and in a time-frame rapid enough to prevent repair of sublethal damage
during the dose delivery process? This is simply a matter of user interface design. Linac delivery
systems have been designed that either duplicate or surpass the accuracy of Leksell Gamma
Knife beam delivery.1 These systems have been designed so that the dose to each isocenter can
be delivered in a time frame that is comparable, if not equivalent, to that of a freshly loaded
Leksell Gamma Knife.
In summary, if tissue differentiation and spatial location are not issues, if dose distributions and
dose gradients can be made equivalent, and if treatment delivery can be as accurate and rapid,
then where lies the debate between a linac and the Leksell Gamma Knife? Can it be the that linac
radiosurgery systems have a greater potential for incorporation of new dose delivery techniques?
Rebuttal
I fully agree with Dr. Goetsch in crediting Lars Leksell and Borge Larsson with the innovative
approach linking stereotactic targeting with multiple nonconvergent beam planning. These tools
have allowed radiosurgeons to deliver large single-fraction doses with submillimeter accuracy. I
also agree with Dr. Goetsch's claims for the required accuracy and convenience. As for his claims
for uptime, all I can say is that at the University of Florida we have treated over 1700
radiosurgery patients and only had to use a backup linac for two patients, giving us an uptime of
99.88% over a thirteen-year period.
While Dr. Goetsch and I agree on most points, there is a single point where we disagree. That
point is not that all linac systems will be as accurate and user friendly as a Leksell Gamma Knife
system, but instead that a linac system can be made to match the accuracy and convenience of a
Leksell Gamma Knife. For this to occur, some modifications or additions to the basic medical
linac are required. If asked, "Can I purchase a stereotactic head frame and, using my standard
external beam planning system and room lasers, deliver high quality radiosurgery?," my reply
would be "no." But if a facility is willing to obtain planning code tailored specifically to
radiosurgery planning, and to adapt their linac to provide the accuracy, stability, and ease of use
required for the small field, high precision, single fraction therapy, then my reply would be "yes."
In his statement Dr. Goetsch appears to make the assumption that a radiosurgery patient must be
planned and treated on the same day. Currently we obtain nonstereotactic MR scans of our
135
patients a day or two prior to the planned radiosurgery. We develop a plan for that patient based
on the nonstereotactic MR scan. On the day of the radiosurgery, we apply the stereotactic head
ring and obtain a CT scan. Fusing the previously nonstereotactic MR scan to the stereotactic CT
scan provides the transform required for plan transfer. On the radiosurgery day, it only requires
10–15 minutes for image fusion, plan transfer and plan verification. We know the time required
for treatment, and can adjust the treatment schedule accordingly.
As for current and future innovations I find the advent of mMLC for high resolution intensity-
modulated planning, extracranial stereotactic approaches, fractionated stereotactic radiotherapy
systems and high dose-rate linacs to more than compensate for the Leksell Gamma Knife's
inclusion of limited automatic isocenter positioning.
AGAINST THE PROPOSITION: Steven Goetsch, Ph.D.
Opening Statement
The first Gamma Knife was created in 1968 by Lars Leksell and Borge Larsson. The device
expanded on a number of previous concepts, including x-ray and proton radiosurgery,
stereotaxis, and Co-60 teletherapy. It offered unparalleled spatial accuracy and remarkable
conformity of dose delivery (90%–50% isodose falloff in as little as 1.0 mm). All that was
lacking in 1968 was modern transaxial imaging, such as computed tomography and magnetic
resonance imaging, and high speed computers. The advent of angiography in the mid-1970s
made visualization and treatment of arteriovenous malformations possible, and thousands were
treated in a few years. Invention of the CT scanner by Hounsfield in 1972 eventually eliminated
the necessity of administering painful, difficult pneumoencephalograms to all Leksell Gamma
Knife patients. This allowed the device to treat brain tumors as well as functional diseases such
as parkinsonism and trigeminal neuralgia for which it was originally designed. Introduction of
the Leksell GammaPlan treatment planning program with a modern UNIX workstation in 1992
made it possible to quickly devise complex, multi-isocenter treatment plans with multiplanar
dose planning. Development of stereotactic MRI allowed unparalleled visualization of target and
normal tissues.
Today the virtues of the Leksell Gamma Knife include remarkable geometric convergence (less
than 0.3 mm for all 201 beams), rapid dose falloff, extremely high reliability and dedication to
cranial applications. Neurosurgeons have frantically busy schedules, including a major role in
trauma cases, and have learned to count on the availability of the Leksell Gamma Knife at
predictable times, typically early in the morning, so that clinical and surgical schedules can be
reliably scheduled. Nondedicated linear accelerator based radiosurgery frequently involves
starting procedures at the end of a long clinical day, with treatments often stretching into the
night hours.
The San Diego Gamma Knife Center treated 1090 patients in its first 62 months of operation with
only two days of planned downtime. No Leksell Gamma Knife cases were postponed or
cancelled because of machine unavailability. A reload of the Model U unit was conducted in
January, 2001. This required only 21 days of clinical downtime (the newer Models B and C
would require about 10–15 days). This yields an overall system availability of 98.3% over 63
months, including reloading. Very few linear accelerators with complex stereotactic radiosurgery
accessories can match this record of availability.
136
The Leksell Gamma Knife has not remained frozen in time. The new Model C with an Automatic
Positioning System offers both computer control of the irradiation process and Record and Verify
capability. The treatment plan is networked to a console which controls the X, Y, and Z
stereotactic coordinates and can move the patient through a limited range of motion between
shots. The treatment time, X, Y, and Z positions, helmet size, gamma angle and shot number are
all automatically checked. This promises to decrease the reported misadministration rate even
more. I believe this new innovation will continue to assure the place of the Leksell Gamma Knife
as the gold standard of radiosurgery.
Rebuttal
The Leksell Gamma Knife offers an extraordinary number of "beam ports" providing a
tremendously conformal absorbed dose distribution. This has been true since the first prototype
33 years ago. What has changed since then is that the "all or nothing" irradiation schemes first
used with single-shot plans have been replaced by modern multi-shot plans utilizing differing
helmet sizes, shot weights and patient head angles. Up to 100 of the 201 gamma ray portals can
also be selectively plugged. The planning challenge is to effectively harness all of these myriad
possibilities. The Spring, 2001 issue of Business Week magazine recognized the University of
Maryland for use of "data mining" to create Leksell Gamma Knife treatment plans in its annual
"Masters of Innovation" awards. Thus, everything old is new again.
The original Gamma Knife was devised to treat functional disorders such as trigeminal neuralgia
and Parkinson's disease. The modern Leksell Gamma Knife is still unsurpassed in treating single-
shot 4 mm diameter volumes very rapidly and efficiently, with remarkably little scattered dose to
adjacent critical structures. This is still very difficult to do with linac-based radiosurgery.
Brilliant innovators such as Professor Bova and his neurosurgical colleague Bill Friedman have
achieved outstanding results in accelerator-based radiosurgery. Precise figures for the total
number of linac radiosurgery treatments performed annually are not available, but the number is
clearly growing as the technique gains acceptance. Still, the 65 American Leksell Gamma Knife
Centers treated over 10 000 patients in calendar year 1999, clearly a large fraction of all
radiosurgery performed in this country. With 143 Leksell Gamma Knife units installed
worldwide and over 150 000 patients treated, this device is certainly not obsolete today, and will
not become obsolete for many years to come.
REFERENCE
1. W. A. Friedman and F. J. Bova, "The University of Florida radiosurgery system," Surg.

Neurol. 32, 334–342 (1989).
137
2.17. In stereotactic radiosurgery, implanted fiducials are superior

to an external coordinate system
Eric G. Hendee and Wolfgang A. Tomé

OVERVIEW
Successful stereotactic radiosurgery requires the delivery of radiation dose exactly to the volume
of tissue to be treated. One approach to accomplishing exact delivery is to immobilize the patient
within an external coordinate system. Another approach is to use implanted fiducial markers for
patient alignment. Both approaches have their advocates, and their comparative merits are
discussed in this Point/Counterpoint.
Arguing for the proposition is Eric G. Hendee, MS. Mr. Hendee completed his graduate and
clinical training at the University of Wisconsin Madison, where he served as a clinical physicist
and Assistant Professor through the University of Wisconsin LaCrosse. He is currently practicing
at Waukesha Memorial Hospital in Wisconsin and is certified by the American Board of
Radiology in Therapeutic Radiological Physics. He has extensive experience in commissioning,
treatment planning, and training for stereotactic radiosurgery. He also developed the UW ADCL
program for gamma knife quality assurance, and has worked in the medical device industry to
develop quality assurance tools for radiation oncology.
Arguing against the proposition is Wolfgang Tomé, Ph.D. Dr. Tomé obtained his Ph.D. in
theoretical physics from the University of Florida and completed a two year residency in
Radiation Oncology Physics at the Shands Cancer Center at the University of Florida. He is
currently working as an Assistant Professor of Human Oncology at the University of Wisconsin
and is board certified by the American Board of Radiology in Therapeutic Radiological Physics.
His fields of primary expertise include Stereotactic Radiosurgery, Frameless Optically Guided
Fractionated Stereotactic Radiotherapy and Intensity Modulated Radiotherapy, as well as
Ultrasound Guided Stereotactic Extracranial Radiotherapy.
FOR THE PROPOSITION: Eric G. Hendee, M.S.
Opening Statement
The use of fixed points of reference, or fiducials, has its origin in ancient celestial navigation as a
straightforward and effective method of localization. In stereotactic radiosurgery, the use of
implanted fiducials is efficient, convenient, and cost effective, with proven accuracy comparable
to frame-based systems. In addition, implanted fiducials provide patient verification from portal
images obtained at the time of treatment.
There are four general methods to define a stereotactic coordinate system: (1) a frame rigidly
attached to the skull, (2) a temporary frame that conforms to the patient (so called relocatable
frames), (3) fixation of fiducials to the patient that are referenced to the treatment target, and (4)
the use of anatomical surfaces, lines or points that are defined with respect to the target. The first
138
two methods are referred to as "frame-based," the last two are termed "frame-less." Frame-based
systems may use the frame for both patient immobilization and localization, whereas frame-less
systems allow the design of rigid patient immobilization independent of localization.
The concept of stereotaxis was developed around frame-based systems for neurosurgery early in
the last century, and has evolved to the widespread use of frame-less methods. The use of
implanted fiducials for high precision radiotherapy and radiosurgery is a more recent
development, first appearing in the literature 10–12 years ago.
Positional accuracy of implanted fiducials has been extensively investigated, including

comparison with the accuracy of frame-based systems. Ultimately, the characteristics of the
imaging procedure determine set-up accuracy for either system, but it is fair to state that in
careful hands there is little, if any, difference in the accuracy obtainable with frameless or frame-
based systems. The fundamental advantage of implanted fiducials over external reference frames
is that the stereotactic coordinate system is permanent. Therefore, the target can be localized with
respect to the treatment beam based on the imaged location of the fiducials. The image provides
information on the patient and isocenter setup, whereas an external reference system provides
information only on the external coordinate system, requiring one to assume that the target is
always in the same relative position. From a quality-control standpoint, it is helpful to remove
intermediate steps in localizing the stereotactic frame space with respect to the linac isocenter.
The use of frame-based systems involves frame placement, imaging, treatment planning, QC, and
treatment, often making the day long for the patient and staff. By comparison, implanted fiducials
are quicker and reduce the chance of error caused by an exhausted staff treating late in the
evening. Implanted fiducials also reduce the chance of compromising the quality of the treatment
plan due to time constraints.
While this debate focuses on radiosurgery, it is undeniable that implanted fiducials facilitate
stereotactic radiotherapy (i.e., fractionated) and treatment of recurrent disease. This is because
they are permanent and therefore always available for precise localization. Finally, implanted
fiducials can be used to localize and treat extracranial lesions, particularly when implanted in a
mobile target such as the prostate. With the advent of improved portal imaging systems that can
distinguish fiducials in real time, it is likely that extracranial applications will increase.
Rebuttal
Dr. Tomé's position is that an invasive fixation system is more accurate in immobilizing the
patient, and I agree with him in that regard. However, in addition to immobilization, stereotactic
radiosurgery and radiotherapy have many components which contribute to the end result,
including localization, fractionation, implementation, and verification.
Several studies have demonstrated comparable localization accuracy between frame-based and
frameless systems. Also, localization of the frame is not the same as localization of the patient. In
other words, frame-based systems provide no verification that the frame has not moved after the
imaging study or that there is any flexure associated with positioning. On the other hand,
implanted fiducials cannot move relative to the patient. The margin of 1 mm in Dr. Tomé's
opening statement is questionable, and is applicable for frame-based treatments as well since
there is no verification of target position at time of treatment. Also, margins are subject to
limitations introduced by the finite number of cones, which often occur in steps of 5 mm.
139
Two important aspects of patient treatment are convenience and dignity. Placement of an
invasive head ring is uncomfortable for the patient, and does not allow fractionation. The head
ring also requires imaging and treatment on the same day, a restriction that does not exist for
implanted fiducials.
For CT imaging, there should always be a rigorous QC program in place, as required for frame-
based systems and external-beam patients.
With regard to verification, triangulation with fiducials allows the isocenter to be correctly
positioned, even if the patient is not in exactly the same position as the imaging study. Values for
yaw, pitch and roll are reported and the clinical team establishes acceptable limits, but this does
not preclude positioning the target at the isocenter. Still, an effective immobilization system must
be in place to keep the patient from moving during treatment.
In conclusion, after considering the entire stereotactic program, one can identify a number of
advantages of implanted fiducials over frame-based systems. We are obliged to remain current
with the technologies that we adapt for our patients. As neurosurgery is progressing with
frameless systems, we should follow these advances closely for possible adaptation to
radiotherapy.
AGAINST THE PROPOSITION: Wolfgang A. Tomé, Ph.D.
Opening Statement
For stereotactic radiosurgery (SRS) and especially for functional stereotactic radiosurgery
(functional-SRS), the relationship of the collimated x-ray beam to a specific target within the
treatment room is critical. The success of radiosurgery lies in the fact that the patient's position is
fixed in space with an invasive fixation system (stereotactic head ring) throughout the imaging
and treatment process, i.e., the patient is treated in the same position as imaged. With such
fixation, only uncertainties caused by limitations in imaging, and mechanical uncertainties in the
radiosurgery system, need to be accounted for during treatment planning and delivery. Hence, a
high degree of conformality can be achieved while simultaneously sparing normal tissues.
A noninvasive immobilization system coupled with the use of implanted fiducial markers
introduces additional uncertainties in treatment planning and delivery due to patient repositioning
and imaging of implanted fiducials. These additional uncertainties have to be accommodated in
the planning and delivery process, and lead to increased margins of uncertainty around the
treatment volume. This uncertainty results in inferior conformality when compared with invasive
fixation systems. To put it more plainly, the margins around the target tissue have to be
increased, so more normal tissue is included in the treatment volume. This is not a trivial point as
the following example shows.
For a spherical 30 mm lesion, an additional inaccuracy of 1 mm due to imaging and localization

of the implanted fiducial system implies that an extra margin of at least 1 mm must be considered
to exist around the lesion. This margin, in turn yields an increase in irradiated normal tissue of up
to 21%. While this increase may be acceptable for patients with malignant lesions or brain
metastases because of the inherent biological uncertainties at the edges of these lesions, it is
clearly inappropriate for benign lesions in which the interface between lesion and normal tissue is
very well defined.
140
The most commonly-used external stereotactic localization system, the Brown–Roberts–Wells

(BRW) System, is an independent localization method. It consists of nine carbon-fiber rods
arranged into three N-shaped structures placed at 120-degree angles around the patient's head.
Three N-shaped fiducials provide an overdetermined system of linear equations that allow for the
unique definition of an independent, absolute coordinate system. Therefore, one does not have to
verify the table indexing of the CT scanner, since the slice position of each scan can be uniquely
determined. This may or may not be the case for an implanted fiducial system, which therefore
requires a rigorous QC program for the CT scanner.
It follows from the above discussion that implanted fiducial systems are not superior to an
external coordinate system for SRS. Rather, they are at best equivalent, and may well be inferior.
Rebuttal
As pointed out by my colleague Eric Hendee, the use of implanted fiducials establishes the
imaging procedure as the principal influence on setup accuracy at the time of treatment. This
uncertainty is not present in frame-based radiosurgery. Therefore, implanted fiducial systems
cannot be superior to frame-based fiducial systems in terms of accuracy. In stereotactic
radiosurgery, we aim to conform the prescription dose shell as close as possible to the target—
i.e., to provide optimal conformal radiotherapy. In this situation, cost effectiveness of a
procedure cannot be used as an argument for superiority if the outcome is inferior in terms of
accuracy. Since frame-based radiosurgery offers a higher degree of accuracy, it is clearly the
method of choice for optimal stereotactic radiosurgery.
Implanted fiducials can be employed in fractionated stereotactic radiotherapy and extracranial

stereotactic radiotherapy. Their use in these applications, however, is still affected by the
uncertainty of localizing the fiducials by using orthogonal radiographic images. Translations of a
target can be corrected rather easily using orthogonal radiographic images and implanted
fiducials. However, it is very difficult to identify rotational misalignment of a target using this
imaging technique. Therefore, it stands to reason that minimally-invasive Optically Guided
Patient Localization Systems, which correct for both translational and rotational misalignment of
the target, as well as ultrasound-guided systems, are preferable since they permit interfraction
patient localization that is equivalent to frame-based radiosurgery1 and CT-guided extracranial
radiotherapy.2
REFERENCES
1. S. L. Meeks, F. J. Bova, T. H. Wagner, J. M. Buatti, W. A. Friedman, and K. D. Foote,

"Image localization for frameless stereotactic radiotherapy," Int. J. Radiat. Oncol., Biol., Phys.
46, 1291–1299 (2000).
2. J. Lattanzi, S. NcNeeley, A. Hanlon, T. E. Schultheiss, and G. E. Hanks, "Ultrasound-
based stereotactic guidance of precision conformal external beam radiation therapy in clinically
localized prostate cancer," Urology 55, 73–78 (2000).
141
2.18. Three-dimensional rotational angiography (3DRA) adds

substantial information to radiosurgery treatment planning of
AVM'S compared to angio-CT and angio-MR
Carlo Cavedon and Frank Bova

OVERVIEW
Three-dimensional rotational angiography (3DRA) is available in most institutions where
arteriovenous malformations (AVMs) are treated radiosurgically. Some physicists believe that
3DRA should be employed to guide AVM treatment with radiosurgery. Other physicists feel that
computed tomographic angiography (CTA) is the preferred technique for use with radiosurgical
treatment of AVMs. This difference in opinion is debated in this month's Point/Counterpoint.
Arguing for the Proposition is Carlo Cavedon, Ph.D. Dr. Cavedon graduated with a Ph.D. in
physics in 1992 from Padova University, Italy. His research field was experimental nuclear
physics. He then specialized in medical physics at the University of Milano. He has been serving
as a medical physicist in Vicenza Hospital since 1997, where he conducted research on dosimetry
and monte carlo simulations. His current research interests are robotic radiosurgery, functional
imaging, and image processing aimed at treatment planning. He is a member of the AAPM and
the Italian Association of Medical Physics. He has had several teaching appointments in Italian
medical physics courses.
Arguing against the Proposition is Frank J. Bova, Ph.D. Dr. Bova received B.S. and M.S. degrees
in Biomedical Engineering from Renesslear Polytechnic Institute, and a Ph.D. in Nuclear
Engineering Sciences with specialization in Medical Physics from the University of Florida. Dr.
Bova joined the Department of Radiation Oncology at the University of Florida in 1978. In 1991
he was appointed Einstein Fund Professor of Computer-Assisted Stereotactic Neurosurgery in the
Department of Neurosurgery where he currently holds the rank of Professor. Dr. Bova also holds
appointments in the Department of Nuclear and Radiological Sciences, Department of
Biomedical Engineering and the Department of Neuroscience. He is a Fellow of the American
Association of Physicists in Medicine, the American College of Radiology and the American
Institute of Medical and Biological Engineering. Dr. Bova holds seven patents in stereotactic
guidance.
FOR THE PROPOSITION: Carlo Cavedon, Ph.D.
Opening Statement
Angiographic images are used extensively in planning radiosurgery for arteriovenous

malformations (AVMs). Visual superposition of angiographic images onto CT slices is not
possible because CT is a tomographic visualization process, whereas angiography is a projection
imaging process. This difference leads to a lack of visual correlation between the two techniques.
Still, there is a need for image correlation between the two methods. This need can be met by
fully exploiting the capabilities of three-dimensional rotational angiography (3DRA), a
142
technology that has only recently become available. Though used mainly for 3D visualization,
3DRA allows tomographic reconstruction so that datasets can be spatially registered to CT
images, as well as to other tomographic modalities such as MRI. Spatial accuracy of the
registration can be a concern, but studies have confirmed its suitability for clinical applications. 1,2
Advantages of 3DRA in planning interventional radiology have been described. 3,4,5 Many of
these advantages also apply to radiosurgery planning, even if they are related solely to 3D
visualization rather than tomographic use of 3DRA data. Other modalities such as CT
angiography (CTA) and MR angiography (MRA) offer tomographic information valuable for
AVM radiosurgery planning.6,7 These techniques are well known and more available than 3DRA,
but they lack important characteristics such as sufficient speed (allowing use of a smaller amount
of contrast medium) and selectivity (offering the possibility of injecting contrast medium
separately into different feeders (e.g., carotid and vertebral arteries) to reveal information about
the AVM structure).8
CT and 3DRA are both x-ray transmission modalities, which provide an advantage over MRA
for which a multimodality registration would be required. This advantage is of key importance
when using frameless radiosurgery techniques, where 3DRA is of maximum benefit because
image registration allows correlation of angiography to CT without the need for a stereotactic
frame.
At present, 3DRA equipment is not available in every radiosurgery center. 3DRA is a significant
imaging improvement, however, and will undoubtedly become widely available in the near
future. For the time being, 3DRA examinations with frameless operation can be performed in an
imaging center, and the data used (even at a different time) in a separate radiosurgery center. A
survey among major 3DRA manufacturers reveals that the distribution of 3DRA equipment is
probably wider than imagined. In Europe, for example, there are approximately the same number
of 3DRA machines as there are radiosurgery centers, although the two are not necessarily in the
same locations. It is likely that available 3DRA equipment within the community simply needs to
be located. Physicists must then select the appropriate registration technique and perform
commissioning tests. This is not an easy job, but it is certainly worth doing.
Rebuttal
I share Dr. Bova's healthy skepticism of unsound methods, and agree with his criticisms of fixed
projection angiography. But 3D rotational angiography is a true 3-dimensional technique and its
novelty, compared to orthogonal projections, lies in its ability to surmount the limitations of
fixed-position angiography. A key advantage of 3DRA is that the number of projections is not
two (or three), but on the order of one hundred. This allows thin slice reconstruction to be
performed, with the typical case of 2563 voxels and an 18 cm field of view for a slice thickness of
0.7 mm. The main difference of 3DRA compared with CT is the use of a cone beam instead of a
fan beam. This should certainly not be regarded as a disadvantage, since cone beam tomographic
reconstruction is used in high-level algorithms. Indeed, these algorithms are now beginning to be
used in new-generation CT equipment.9,10
If a flaw can be identified in 3DRA, it may be with regard to its timing characteristics, which are
as yet not comparable to fixed projections. This prevents the use of 3DRA for distinguishing
between feeders and draining vessels. Nevertheless, as outlined in my opening statement, 3DRA
has significant other advantages which are useful in angiographic examinations.
143
Dr. Bova starts his proposition by praising methods based on numbers. I certainly agree with him,
as demonstrated by the case under discussion. Registering 3DRA to CT by means of a state-of-
the-art algorithm, e.g., mutual information maximization, provides a quantitative, operator-
independent method to exploit this new technique in AVM radiosurgery.
Finally, physicists should be keen to explore new capabilities rather than limit themselves to the
application of consolidated ones. 3DRA represents an innovation whose potential, though still to
be explored in all of its aspects, can elicit significant information of vital use in radiosurgery of
vascular malformations.
AGAINST THE PROPOSITION: Frank Bova, Ph.D.
Opening Statement
As medical physicists, we must be certain that we clearly see the underlying principles of our
science. As Lord Kelvin once said, "If you can not express your knowledge in numbers, it is of a
meager and unsatisfactory kind." For many decades, medical physicists have relied upon
orthogonal radiographs to define the geometry of radioactive implants. During the 1960s and
1970s, the reconstruction debate concerned stereo-shift radiography versus orthogonal
radiography. Our long success with orthogonal reconstructions of implants may have blinded us
to the underlying requirements of orthogonal spatial definition. We may have forgotten that the
first requirement of orthogonal reconstruction is identification of identical points in each view.
As implants became more complex, and more and smaller sources were added, this requirement
posed more of a problem. Undaunted by this warning, the high spatial accuracy promised by
orthogonal reconstruction continued to seduce medical physicists into trying to fit this square peg
into new round holes. Before CT, when stereotactic systems were first introduced into
neurosurgical procedures, stereotactic targets often were inferred from points identified on
orthogonal ventriculograms. While these techniques provided some measure of success, the
advent of true 3D datasets provided by CT scanning, and subsequently MR imaging, have
rendered them obsolete. As treatment planning increased in complexity, precise complex 3D
target descriptions were required. These descriptions are best supplied by a CT or MR scanner.
The potential errors of orthogonal imaging of a solid structure can be easily demonstrated.
Imagine an ellipse with its center at the origin and its long axis oriented at 45 degrees to the XY
axis. The orthogonal projection of this ellipse to either axis can be encompassed by a circle with
a diameter of the length of the ellipse divided by the square root of two. Unless unique points on
the target's extreme edges can be identified in each view, it is impossible to deduce the original
length of the long axis of the ellipse. Instead of an ellipse, suppose you placed the letter "C" at
the same point in space. Orthogonal projections would not yield the shape of this target, and an
accurate description of the target would be impossible.
The strength of orthogonal projections is their ability to accurately reconstruct points in space.
When we reconstruct an implant, we are reconstructing identifiable points. When we are defining
stereotactic targets, however, we are defining 3-dimensional volumes in which individual points
are seldom identifiable. While it has often been suggested that a third view can provide the
information required to define a specific object through orthogonal imaging, it can easily be
demonstrated that the new solution can be fooled by a simple rotation of the target to render the
reconstructed target invalid. The only method of providing a true description of a 3-dimensional
object is to image the object with a true 3-dimensional imaging technique. Thin slice CT and MR
144
imaging provide such descriptions of clinical targets, and therefore are the image modalities that
should be used when developing conformal treatment plans.
Rebuttal
One of the basic rules of image guided therapy is never say "no" to a high fidelity image detailing
the target tissues. There is no doubt that in selected situations, the addition of 3DRA sheds light
on target definition. At the end of the day, however, the clinical team must choose which
stereotactic coordinates to target and which stereotactic coordinates to spare. The true 3D data set
provided by multislice CTA provides the gold standard for such targeting.
The Achilles heal of 3DRA is the time required to complete the rotation needed for image
acquisition. One should not confuse the time required for a planar angiographic image
acquisition at many images per second, with the time required to complete the rotational needed
for 3DRA.
CT scanners can acquire 16 slices per second. These images can cover the entire transaxial extent
of the target tissue and 8 to 16 mm of axial extent, while maintaining submillimeter pixel
resolution. If 3D prospective imaging is needed, post processing can provide high fidelity
renderings of the target vessels.
While planar angiography remains the gold standard for diagnosis of vascular abnormalities, the
playing field for radiosurgery targeting is firmly grounded in true 3D data sets. Although one can
detect a target in 2D, you need an accurate 3D fix to target from multiple trajectories.
REFERENCES
1. R. R. Bridcut, R. J. Winder, A. Workman, and P. Flynn, "Assessment of distortion in a

three-dimensional rotational angiography system," Br. J. Radiol. 75, 266–270 (2002).
2. F. Colombo, C. Cavedon, P. Francescon, L. Casentini, U. Fornezza, L. Castellan, F.
Causin, and S. Perini, "Three-dimensional angiography for radiosurgical treatment planning for
arteriovenous malformations," J. Neurosurg. 98, 536–543 (2003).
3. R. Anxionnat, S. Bracard, J. Macho, E. Da Costa, R. Vaillant, L. Launay, Y. Trousset, R.
Romeas, and L. Picard, "3D angiography. Clinical interest. First applications in interventional
neuroradiology," J. Neurosurg. 25, 251–262 (1998).
4. T. Abe, M. Hirohata, N. Tanaka, Y. Uchiyama, K. Kojima, K. Fujimoto, A. M. Norbash,
and N. Hayabuchi, "Clinical benefits of rotational 3D angiography in endovascular treatment of
ruptured cerebral aneurysm," AJNR Am. J. Neuroradiol. 23, 686–688 (2002).
5. C. J. Prestigiacomo, Y. Niimi, A. Setton, and A. Berenstein, "Three-dimensional
rotational spinal angiography in the evaluation and treatment of vascular malformations," AJNR
Am. J. Neuroradiol. 24, 1429–1435 (2003).
6. S. Aoki, Y. Sasaki, T. Machida, N. Hayashi, I. Shirouzu, T. Ohkubo, A. Terahara, Y.
Sasaki, S. Kawamoto, T. Araki, and T. Maehara, "3D-CT angiography of cerebral arteriovenous
malformations," Radiat. Med. 16, 263–271 (1998).
7. G. Bednarz, B. Downes, M. Werner-Wasik, and R. H. Rosenwasser, "Combining
stereotactic angiography and 3D time-of-flight magnetic resonance angiography in treatment
planning for arteriovenous malformation radiosurgery," Int. J. Radiat. Oncol., Biol., Phys. 46,
1149–1154 (2000).
8. Y. Kakizawa, H. Nagashima, F. Oya, K. Ito, Y. Tanaka, K. Hongo, and S. Kobayashi,
"Compartments in arteriovenous malformation nidi demonstrated with rotational three-
145
dimensional digital subtraction angiography by using selective microcatheterization. Report of

three cases," J. Neurosurg. 96, 770–774 (2002).
9. S. Bonnet, S. F. Peyrin, F. Turjman, and R. Prost, "Nonseparable wavelet-based cone-
beam reconstruction in 3-D rotational angiography," IEEE Trans. Med. Imaging 22, 360–367
(2003).
10. I. Hein, I. K. Taguchi, M. D. Silver, M. Kazama, and I. Mori, "Feldkamp-based cone-
beam reconstruction for gantry-tilted helical multislice CT," Med. Phys. 30, 3233–3242 (2003).
146
2.19. New methods for precision radiation therapy exceed biological

and clinical knowledge and institutional resources needed for
implementation
Sarah S. Donaldson and Arthur L. Boyer

OVERVIEW
New methods for precision radiation therapy (e.g., conformal, intensity-modulated and
tomographic radiotherapy) have the potential to distribute radiation to tumors and normal tissue
to any desired degree of exactness. Whether such methods are justified in light of the biological
and clinical uncertainty about tumor anatomy and physiology is debatable. Also debatable is
whether precision methods require more time and personnel than can be allocated in a busy
clinic. These debatable issues are the subject of this Point/Counterpoint.
Arguing for the Proposition is Sarah S. Donaldson, M.D. Dr. Donaldson obtained her M.D. from
Harvard Medical School. She completed Radiation Therapy Residency Training at Stanford,
where she is now Professor of Radiation Oncology. She serves as Associate Chair of the
Radiation Oncology Department and Associate Director of the Clinical Cancer Center. Dr.
Donaldson‘s professional interests are in Pediatric Radiation Oncology, and late effects from
cancer treatment. As a clinician, researcher and educator, she carries responsibility for setting
standards for the safe and effective delivery of radiation. She is a past President of ASTRO and
the ABR, and is an elected member of the Institute of Medicine.
Arguing against the Proposition is Arthur L. Boyer, Ph.D. Dr. Boyer is a tenured Professor
(Radiation Physics) at Stanford University School of Medicine. He serves as Director of the
Radiation Physics Division of the Department of Radiation Oncology. He received a B.A. in
Physics from the University of Dallas in 1966, and a Ph.D. in Physics from Rice University in
1971. He has held appointments at the Massachusetts General Hospital and the Harvard Medical
School in Boston, the Cancer Therapy and Research Center in San Antonio, and the M. D.
Anderson Cancer Center in Houston.
FOR THE PROPOSITION: Sarah Donaldson, M.D.
Opening Statement
Among the new advances for precision treatment, intensity-modulated radiation therapy (IMRT)
currently occupies the greatest clinical attention; therefore, I have chosen it for the focus of my
commentary. Aggressive IMRT reporting and marketing promise to improve upon three-
dimensional conformal therapy (3-D CRT) and to yield substantial improvements over existing
technologies.1 Before accepting such assertions, clinicians need to confirm that the promise of
projected dose escalations and organ-sparing techniques will result in improved local/ regional
control and reduced long-term morbidity. Concern with this technology include uncertainties
related to:
147
(1) technical issues of patient immobilization and uncontrolled organ motion;

(2) imaging inaccuracies in determining accurate tumor volume;
(3) lengthy planning and delivery times;
(4) dose inhomogeneities and Dose-Volume Histogram (DVH) interpretation;
(5) quality assurance procedures, verification and documentation of beam delivery;
(6) biologic unknowns including carcinogenesis; and
(7) economic considerations.
Despite the theoretical arguments for incremental benefits of IMRT over standard or 3-D CRT,
there are little clinical data to support general noninvestigative use. Efficacy and complication
data come from very few centers, in which only select adult tumor types have been studied.
Long-term follow-up is not yet available.
Clinicians experienced with IMRT appreciate the inaccuracies associated with masks, moulds,
shells, cradles, vacuum bags, and rectal balloons for patient immobilization. Organ motion
remains beyond physician/physicist control. Dose inhomogeneity is substantial. Thorough
understanding of partial organ tolerance and acceptable DVH values is lacking. Acceptable
quality assurance procedures for IMRT are in their infancy. The need for retreatment following
marginal recurrence has not been addressed. Unique and individualized treatment plans create
long delivery times that are often difficult for the patient, and are disruptive in a busy clinical
department. Both dynamic leaf movement and the step-and-shoot methods take longer than
conventional 3-D CRT. Furthermore, while for some anatomic sites a limited number of standard
IMRT fields may be routine, large tumor and nodal volumes, and doses to diverse areas of the
body in pediatric cancer, cannot follow standard templates. They require even more lengthy
planning and delivery.
Fusion of CT/MRI images is imprecise; the most knowledgeable experts are uncertain about
exact tumor boundaries, thus undermining the scope of precision therapy. The impact of
multifield, low-dose radiation exposure, and higher total body doses from leakage radiation
associated with longer ‗‗beam-on‘‘ times and leaf transmission, carry risks of radiation-induced
carcinogenesis that cannot be accurately assessed. 2
Start-up and maintenance of IMRT equipment are extraordinarily expensive. Additionally, there
is a costly learning curve for professional and technical staff, contributing to an increase in time
per patient for the complexities of IMRT. Despite these higher costs, reimbursement for 3-D
CRT and IMRT is equivalent from most providers. Specific IMRT CPT codes are not yet
developed, and are unlikely to ever fully reimburse actual expenses.
Thus, we must conclude from clinical, biologic and economic standpoints, that IMRT is only
appropriate at a very few highly sophisticated research centers with enormous institutional,
professional and technical commitment. IMRT must remain an investigative tool until clinical
trials comparing IMRT with rival treatment approaches show safety and efficacy using clinically
important end points of tumor control, side effects, quality of life, and health costs. We should
not implement this technology outside a research setting without the answers to these unknowns.
Rebuttal
The new technology of precision radiation therapy ushers in issues beyond those covered in the
definitions of ‗‗biological knowledge‘‘ and ‗‗clinical knowledge‘‘ addressed by Dr. Boyer.
Among the most critical issues are the time, effort, and institutional resources needed to support
148
the new technology, and the need to balance these resources against existing institutional
requirements. It must be recognized that the added efforts to utilize the technology efficiently for
precision radiation therapy need to be integrated into a departmental strategic plan without
adversely impacting the current full-time requirements inherent in the practice of clinical
medicine. This is a challenging task that requires increased personnel and added expense.
Furthermore we must recognize and respond to the many uncertainties that technologic advances
create, because these uncertainties may impact clinical outcome, including localregional control.
Dr. Boyer and I agree that radiation oncologists today must ‗‗explore the potentials and
limitations of new technologies.‘‘ This statement reiterates my position for investigation and
research prior to widespread implementation of precision radiation therapy.
We must conclude from current clinical, biological, and economic standpoints that new methods
for precision radiation therapy belong in a research setting where tumor control, complications,
quality of life, and health costs can be investigated in order to justify a broader implementation
of the new technology.
AGAINST THE PROPOSITION: Arthur L. Boyer, Ph.D.
Opening Statement
New methods for precision radiation therapy are emerging in radiation oncology. Image-guided
prostate seed implants, high dose-rate brachytherapy, and stereotactic, 3-D conformal, and
intensity-modulated radiotherapy are among these new methods. The proposition concerning
these new methods is in fact three separate but interrelated propositions that deal with biological
knowledge, clinical knowledge and institutional resources. I will address the three manifestations
of the proposition in turn.
‗‗Biological knowledge‘‘ is the ability to model and predict the response of the whole patient to a
given course of treatment. I concede that such knowledge is not available for the new methods.
To make accurate predictions one needs to know the genetic profile of the patient and understand
the signal transduction pathways that determine the rate of apoptosis following a course of
treatment. But, models with this level of predictive power are not available for conventional
methods either. A broader interpretation of ‗‗biological knowledge‘‘ is the ability to predict
trends at approximate dose levels based primarily on clinical experience coupled with an
experimental understanding of underlying but unquantified processes. With few exceptions,
conventional therapeutic methods were developed gradually over the years, simultaneously with
the accumulation of this ‗‗biological knowledge.‘‘ I see no different way forward with the new
methods. The lack of biological knowledge did not block techniques used now, and should not
block the implementation of improvements on the methods.
‗‗Clinical knowledge‘‘ refers to the skills needed to anticipate the clinical outcome of a given
treatment procedure accurately enough to make decisions that will prevent unacceptable harm.
Physicians using traditional methods of radiation oncology make these decisions routinely with
less information about the target volume and the three-dimensional distribution of dose over
organs at risk than that provided by the new methods. The problem facing radiation oncologists
using the new methods is that they have additional 3-D dose information, and additional control
over the extent and limits of the dose distribution, but do not know how best to use them.
Experienced radiation oncologists have the necessary background to explore the potentials and
149
limitations of the new technologies. It will take some time before academic centers have
acquired the needed experience and can teach new residents how to use new methods. However,
the foundation for acquiring this knowledge is already in place.
With regard to institutional resources, there are more than 2000 radiation oncology centers in the
US. The style of practice and resources in these centers is spread over a spectrum. The spectrum
encompasses academic centers where expectations and resources foster research and
development as well as private practice centers where the focus is on efficiency and finances.
The implementation of new methods depends largely on the location of the facility in this
spectrum. Implementation means the installation, testing, and commissioning of the new
technology which requires an infrastructure of advanced linacs, computer networks, and skilled
radiation oncologists, physicists, dosimetrists, and therapists. Where these resources and
motivation exist, implementation of the new methods has already occurred and will continue to
occur. Acquisition of new oncology technologies is expensive.
Vendors are still recovering their development costs. In time one expects performance to
improve and costs to lower, leading to a spread of the new methods over more of the facility
spectrum. However, the implementation of the new methods should continued to be pursued
energetically in appropriate environments.
Rebuttal
Dr. Donaldson has chosen to focus on IMRT, and concedes that this new radiotherapy
methodology should be implemented in research settings. She addresses a number of important
aspects of implementing IMRT. Not all of her statements are strictly accurate. For example, she
leaves the impression that an IMRT treatment may be many times longer than a comparable
conventional treatment. With adequately engineered delivery systems that are currently
available, extraordinarily long IMRT delivery times are avoided. I agree completely that data
describing the response of normal structures to low doses is important to understanding the limits
of IMRT.
However, the key debatable issue is the preferred scope and speed of implementation of IMRT,
or in fact of any new treatment modality. On the one hand the radiation oncology community
might begin treating patients with IMRT too hastily, and in the process precipitate experiences
that condemn the process. On the other hand, we might approach the problem so conservatively
that we unduly postpone the realization of the advantages of the new technologies. Good
judgment and common sense must be used to strike a balance.
The question becomes whether it is developed as an investigative tool with the intent of adding
to the knowledge base of radiation oncology, or whether it is regarded as a routine clinical
procedure whose outcome is as certain as the outcome of more familiar procedures. I concede
that new methods should be regarded as the former. It is the responsibility of the facility to
approach a given implementation with the care and caution that is due a new treatment method.
Appropriate quality assurance and record keeping are essential.
Clearly there are many unanswered questions about IMRT, stereotactic radiotherapy, high-dose
rate brachytherapy, and endovascular brachytherapy that will require considerable research
before the outcomes of treatment are predictable. The fact remains that these procedures must be
implemented in enough facilities to provide the pragmatic experience needed to answer the
clinical questions.
150
REFERENCES
1. C. Nutting, D. P. Dearnaley, and S. Webb, ‗‗Intensity modulated radiation

therapy: A clinical review,‘‘ Br. J. Radiol. 73, 459–469 (2000).
2. D. Verellen and F. Vanhavere, ‗‗Risk assessment of radiation-induced
malignancies based on whole-body equivalent dose estimates for IMRT.
151
2.20. Kilovoltage imaging is more suitable than megavoltage imaging

for guiding radiation therapy
Lei Xing and Jenghwa Chang
Reproduced from Medical Physics, Vol. 34, No. 12, pp. 4563–4566, December 2007
OVERVIEW
Image-guided radiation therapy involves the use of imaging to delineate the structures of interest,
to plan appropriate treatment fields, and to ensure that treatments are administered as planned. It
is becoming increasingly common to utilize in-room imaging and localization systems for these
purposes and, most often, these involve use of either MV x rays directly from the linear
accelerators used to treat the patients, or kV x rays from auxiliary x-ray machines. It could be
argued that the kV option is superior because kV imaging is bound to provide better quality
images. On the other hand, some might argue that MV systems are preferable because, among
other things, they use identical geometry to that used for treatment and hence provide more
accurate geometrical information. There are many arguments for and against each of these
modalities and this is the topic of this month's Point/Counterpoint.
Arguing for the Proposition is Lei Xing, Ph.D. Dr. Xing earned his Ph.D. in physics from the
Johns Hopkins University and obtained his medical physics training from the University of
Chicago. He is currently an Associate Professor and Chief of Research in the Department of
Radiation Oncology at Stanford University. His major areas of interest are image-guided and
adaptive radiation therapy, treatment planning, and dose optimization, and the application of
PET/CT and other emerging biological imaging modalities in radiation oncology. He currently
serves on the AAPM workgroups on IMRT and Molecular Imaging in Radiation Oncology, and
the ASTRO Collaborative Committee on IGRT, and has served on the Board of Editors of
Medical Physics. He is certified by the American Board of Radiology in Therapeutic
Radiological Physics.
Arguing against the Proposition is Jenghwa Chang, Ph.D. Dr. Chang obtained an MS degree in
Computer & Information Science, and MS and Ph.D. degrees in Electrical Engineering, all from
the Polytechnic University of New York, Brooklyn, NY. He is board certified by the ABR in
Therapeutic Radiological Physics and the ABMP in Radiation Oncology Physics and is currently
an Associate Member of the Medical Physics Department, Memorial Sloan-Kettering Cancer
Center, New York, NY. His major research interests are intensity modulated radiotherapy,
respiratory gating, and various forms of imaging for radiotherapy including cone-beam CT, portal
imaging, functional MRI, and magnetic resonance spectroscopy.
FOR THE PROPOSITION: Lei Xing, Ph.D.
Opening Statement
When talking about onboard imaging for therapeutic guidance, one has a number of options: the
best imaging plus the best therapy, so-so imaging plus the best therapy, not-so-good imaging plus
152
the best therapy,…, so-so imaging plus so-so therapy. The issue here is not which modality is
better for IGRT, rather where we should settle in this long list of choices. The first thoughts one
may have are what defines the best imaging technology and what is the best therapy machine?
Instead of going through a lengthy description, I simply suggest visits to the exhibit halls at the
RSNA and AAPM annual meetings. Current state-of-the-art imaging devices are displayed each
year at the RSNA meeting, but finding an MV CT scanner there is improbable, much like trying
to find a Co-60 machine on the AAPM exhibition floor. If still unconvinced, perhaps comparing
MV and kV films on a light box in your clinic would help, whence it should become transparent
that, compared to its counterpart, kV x rays are better in providing image guidance. All in all, it is
written on the wall as well as in physics books that the tissue needs to absorb a sufficient number
of photons to be seen, and a structure needs to absorb a significantly different number of photons
from its neighbors to be visualized. Compton scattering simply cannot beat the photoelectric
effect in this regard.
Poor soft tissue contrast alone is sufficient for the MV imaging system to shy away. But there is
more. MV imaging delivers more radiation dose to the patient. IGRT is rapidly moving toward
adaptive replanning1,2,3 and/or real-time (or at least reasonably frequent) feedback of anatomical
information during the beam delivery process,4,5 I simply cannot see how MV alone can meet the
increased demand for frequent imaging. Image truncation in current MV volumetric imaging may
present another problem since volumetric data are required for dose reconstruction6,7 and
adaptive replanning.
Not every system in the real world is made of a combination of the greatest things for various
reasons. Each choice often comes with a different price tag and other pros and cons, and the
optimal choice is a result of balancing different competing factors. In the issue debated here, the
equation in front of us is actually not that complicated. On one hand, one has a kV imaging
system mounted on the LINAC gantry, which is reasonably affordable and provides superior soft
tissue contrast and full 3D anatomic information. On the other hand, there is an MV-based
approach, which uses the treatment beam for volumetric imaging and has a long list of serious
compromises. While the configuration of the MV system is simpler, the sacrifice we have to
make in image quality and patient radiation dose is simply too large to justify the routine use of
this modality in the clinic.
Let me conclude by saying that an ideal IGRT solution should be composed of the best imaging
plus the best therapy machine, and an onboard kV imaging system fits this philosophy well. No
seeing, no hitting. Futuristically, I also see more hope for kV image-guidance since much more
research is being devoted to kV x-ray imaging, which may further enhance the performance of
kV devices. Some day, we may see onboard phase-contrast CT, or inverse-geometry CT, or even
multiplexing nanotube-based CT in radiation oncology clinics. In the spirit of ―the best plus the
best,‖ I am also glad to see that the hybrid of MRI and linear accelerator is on the horizon.
AGAINST THE PROPOSITION: Jenghwa Chang, Ph.D.
Opening Statement
Multiple factors, including clinical usefulness, technical complexity, and cost, must be
considered to determine the best combination of technologies for IGRT. It is a general conception
(or misconception) that megavoltage MV imaging can never compete with kV imaging because
image quality and dose are orders-of-magnitude worse. Recent advances in MV cone-beam
153
computed tomography (CBCT), however, suggest that MV imaging may in fact be more suitable
than kV imaging for many IGRT applications.
Recent improvement in MV imagers has enabled them to produce clinically useful images with
acceptable imaging dose.8,9,10 Although the quality of MV images is generally inferior to kV
images, the difference narrows as the patient thickness increases,10 and almost vanishes for
CBCT due to the much higher scatter for kV CBCT compared to MV CBCT.11 The imaging dose
of MV CT/CBCT is slightly higher but the difference is now less than a factor of 2 and MV
imaging dose can be readily included in the treatment planning process to minimize its adverse
effects.12
MV imaging has been successfully used for two-dimensional IGRT applications including
tracking of implanted markers or bony anatomy for target localization and setup correction13 with
an accuracy comparable to kV imaging.14 Observing intrafractional motion in the beam direction
on an MV imager makes more clinical sense than on a kV imager perpendicular to the beam. MV
CT/CBCT systems have also been successfully applied to three-dimensional setup and
verification with excellent accuracy. 8,9,10 MV CT/CBCT imaging is superior in its linear
relationship between relative electron density and CT number for dose calculation. 9 Because
artifacts due to metal objects and beam hardening are less critical for MV sources, MV CBCT
scans have been acquired to complement diagnostic CT scans when these artifacts are severe. 9
Technical complexity is a major concern for clinical implementation of IGRT because more
complex systems demand more quality assurance (QA) and are more susceptible to errors. MV
imaging using the MV source and imager on a linear accelerator is technically simple and robust,
and has a lower cost for hardware than kV imaging that requires an extra detector and x-ray
source. The QA for MV CBCT systems is basically the same as that for an MV imager, while the
more complex QA for kV CBCT systems15 demands additional manpower and therefore costs
more.
Yoo et al.15 pointed out that the most critical QA of an IGRT system involves maintaining
geometric accuracy for patient repositioning. In this regard, kV IGRT systems are overly
complex because three isocenters (MV source, kV source, and laser) need to be identified and
constantly checked.15 MV IGRT systems with coincident treatment and imaging isocenters, on
the other hand, are true ―what you see is what you treat‖ systems that can do without the
complex, error-prone calculation of systematic shifts between isocenters.
In conclusion, clinical usefulness of MV imaging is comparable or in some cases even superior

to kV imaging for IGRT. There are no reasons to buy a more expensive and error-prone system if
a significantly cheaper and notably less technically challenging alternative is available.
Rebuttal: Lei Xing, Ph.D.
In order to debate whether kV or MV imaging is more suitable for IGRT, the clinical goals must
be clear. If IGRT is all about locating static metallic fiducials and bony structures, nothing
discussed here matters, since even MV projection images will do the job. In reality, the drive to
―see‖ the soft tissue and tumor target, mostly in real time, promotes the development of on-
treatment imaging technology. MV image quality is limited by the physics, whereas the problems
with kV imaging pointed out by Dr. Chang can be addressed. For example, metal artifacts with
kV CT imaging can be removed with appropriate reconstruction algorithms. With the ongoing
research in the field, one has every reason to believe that the quality of kV imaging, which is
154
already superior to the competing MV technology, will be further improved. Indeed, with the use
of primary modulation with spatial variant attenuation materials for scatter removal, a significant
improvement in kV CBCT image quality has been demonstrated.16
To be fair, in-line MV imaging is useful in the localization of implanted metallic fiducials.

However, it generally requires a larger marker size and, in a realistic clinical situation, the auto-
detection of the fiducials is more difficult, especially when a high temporal resolution is required.
MV imaging alone is hardly a solution for real-time fiducial tracking and a simultaneous kV
beam seems to be desirable.17
Yes, in principle the large MV imaging dose can be accounted for during treatment planning.
But, ―accounting for‖ does not mean that no extra dose, which can otherwise be avoided by
shaping the treatment beams, will be delivered to the adjacent sensitive structures. For a breast
cancer patient, for example, the imaging dose is delivered not only to the ipsilateral breast, but
also the contralateral breast, the heart, and the lungs. For patients with a long life expectancy
after radiation therapy, radiation dose resulting from real-time and/or routine adaptive imaging is
of particular concern. MV imaging falls short in this aspect and seriously compromises the value
of image guidance.
QA of the kV imaging device entails additional effort, but the task is quite manageable and can
be automated by well-designed phantom and analysis software tools. 18 Given the potential impact
resulting from image-guided 4D and adaptive therapy, the minimal extra QA effort is clearly
worthwhile. After all, 2D/3D kV imaging is providing us with better, and often additional,
information for therapeutic guidance.
Rebuttal: Jenghwa Chang, Ph.D.
What is the best IGRT system? Should it be the best imaging plus the best therapy device as Dr.
Xing proposed, or, as I pointed out, the system that best meets overall clinical needs? For
conventional radiotherapy, the imaging devices for simulation and the therapy devices for
treatment can be separately optimized because the simulation and treatment delivery processes
are independent. For IGRT, however, optimizing each device on its own may not lead to the best
solution because the imaging device is used to guide the therapy device.
Dose, time, and personnel constraints will limit routine use of lengthy, repeated on-board CT
scans for adaptive replanning or real-time feedback of anatomic information. Instead, tracking
implanted markers is often sufficient for monitoring intrafractional tumor motion. Although both
kV and MV imaging can be used,13 tracking marker motion relative to the beam direction with an
MV imager is more intuitive and makes more clinical sense. In cases where the benefits of
replanning outweigh dose and practical time concerns, MV CBCT provides more accurate CT
number information and is less sensitive to metal artifacts.9
Dr. Xing is correct in stating that kV imaging is superior to MV imaging in terms of image
quality and dose. However, the difference is probably clinically insignificant. Major research
efforts have made the soft tissue contrast and imaging dose of MV imaging comparable to those
for kV imaging.8,9,10 In fact, neither kV nor MV on-board imaging devices for IGRT can compete
with diagnostic imaging devices in image quality, but both can be registered with simulation
images for setup and verification. Imaging dose can also be easily included in treatment planning
for both modalities,12 though adding the kV dose requires new beam data that are not collected as
part of a normal commissioning process.
155
In conclusion, we should focus on the overall clinical needs when evaluating IGRT devices. MV
IGRT systems are truly ―no seeing, no hitting‖ and more cost-effective, and therefore more
suitable than their kV counterparts for guiding radiotherapy.
REFERENCES
1. A. de la Zerda, B. Armbruster, and L. Xing, ―Formulating adaptive radiation therapy

planning into closed-loop control framework,‖ Phys. Med. Biol. 52, 4137–4153 (2007).
2. C. Wu et al., ―Re-optimization in adaptive radiotherapy,‖ Phys. Med. Biol. 47, 3181–
3195 (2002).
3. L. Xing et al., ―Overview of image-guided radiation therapy,‖ Med. Dosim. 31, 91–112
(2006). P. J. Keall, V. R. Kini, S. S. Vedam, and R. Mohan, ―Motion adaptive x-ray therapy: A
feasibility study,‖ Phys. Med. Biol. 46, 1–10 (2001).
4. G. C. Sharp et al., ―Tracking errors in a prototype real-time tumor tracking system,‖
Phys. Med. Biol. 49, 5347–5356 (2004).
5. Y. Yang, E. Schreibmann, T. Li, C. Wang, and L. Xing, ―Evaluation of on-board kV cone
beam CT (CBCT)-based dose calculation,‖ Phys. Med. Biol. 52, 685–705 (2007).
6. L. Lee, Q. Le, and L. Xing, ―Retrospective dose reconstruction based on cone beam CT
and MLC-log file,‖ Int. J. Radiat. Oncol. Biol. Phys. (2007) (in press).
7. J. Chang et al., ―Observation of interfractional variations in lung tumor position using
respiratory gated and ungated megavoltage cone-beam computed tomography,‖ Int. J. Radiat.
Oncol. Biol. Phys. 67, 1548–1558 (2007).
8. O. Morin et al., ―Megavoltage cone-beam CT: System description and clinical
applications,‖ Med. Dosim. 31, 51–61 (2006).
9. K. Langen et al., ―TomoTherapy's implementation of image-guided and dose-guided
adaptive radiation therapy,‖ in Proceedings of AAPM Summer School, 2006
10. S. Lim et al., ―An in vivo comparative study of the MV and kV cone beam computed
tomography image quality of a lung patient,‖ Med. Phys. 33, 2221–2221 (2006).
11. K. Sidhu et al., ―Optimization of conformal thoracic radiotherapy using cone-beam CT
imaging for treatment verification,‖ Int. J. Radiat. Oncol. Biol. Phys. 55, 757–767 (2003).
12. M. G. Herman, ―Clinical use of electronic portal imaging,‖ Semin. Radiat. Oncol. 15,
157–167 (2005).
13. L. Pisani et al., ―Setup error in radiotherapy: On-line correction using electronic
kilovoltage and megavoltage radiographs,‖ Int. J. Radiat. Oncol. Biol. Phys. 47, 825–839 (2000).
14. S. Yoo et al., ―A quality assurance program for the on-board imager®,‖ Med. Phys. 33,
4431–4447 (2006).
15. L. Zhu, J. StarLack, N. R. Bennett, T. Li, L. Xing, and R. Fahrig, ―Improved scatter
correction for x-ray conebeam CT using primary modulation,‖ in Physics of Medical Imaging,
edited by J. Hsieh and M. J. Flynn, Proceedings of the SPIE, Vol. 6510 (2007), p. 65101U.
16. R. Wiersma and L. Xing, ―3D Real-time internal fiducial tracking by use of combined kV
and MV imaging,‖ Med. Phys. (submitted).
17. W. Mao, L. Lee, and L. Xing, ―Development of phantom and automated analysis
software for QA of on-board imaging device,‖ Med. Phys. (submitted).
156
CHAPTER 3
Brachytherapy
3.1. HDR Brachytherapy makes Cs-137 intracavitary therapy for

cervix cancer a breach of good practice
Michael T. Gillin and Jatinder R. Palta

Reproduced from Medical Physics, Vol. 26, No. 4, pp. 499-501, April 1999
OVERVIEW
In many institutions high dose-rate (HDR) brachytherapy has become the method of choice in
treating cancers of the cervix and uterus. Advantages include improved dose distributions,
reduced treatment times, and decreased hazards to personnel compared with the older treatment
method employing sealed 137Cs sources. Still, the latter method is used in many institutions, in
part because they have not yet acquired the technology and expertise for HDR brachytherapy.
This edition of the Point/Counterpoint series explores the issue of whether these institutions are
meeting the current standards of good clinical practice.
Arguing for the Proposition is Michael Gillin. Michael T. Gillin, Ph.D., has worked in the
Department of Radiation Oncology at the Medical College of Wisconsin since 1975 and is a
professor of radiation oncology. His graduate degrees are from the University of California,
Davis. He completed a two year fellowship in Radiological Physics at Walter Reed Army
Medical Center. His research interests include brachytherapy and clinical trials. He is chairman
of the RTOG Medical Physics Committee and has helped organize the first RTOG prostate
implant protocol. He is the head of a five person medical physics group and his daily activities
are centered on insuring that patients are treated in a safe and appropriate manner in clinics in
which he and his clinical and physics colleagues practice. Professor Gillin loves stimulating
discussions with his colleagues, his teenage children, and, most of all, his wife.
Arguing against the Proposition is Jatinder Palta. Jatinder R. Palta, Ph.D., is Professor and
Chief of Physics within the Department of Radiation Oncology at the University of Florida. He
received his Ph.D. in medical physics in 1981 from the University of Missouri and completed
his postdoctoral training at the M. D. Anderson Hospital in 1982. Dr. Palta‘s professional
career has developed around the traditional triad of scientific investigation, clinical
contributions, and teaching. His research interests include radiation dosimetry, three-
dimensional treatment planning, and conformal therapy. The goal of his research endeavors is
to bring new, emerging technologies safely into the clinic. He is currently the chairman of the
Radiation Therapy Committee of the American Association of Physicists in Medicine.
157
For the proposition: Michael T. Gillin
Opening Statement
Continuous evolution of radiation oncology techniques and treatments is the norm. The point
in time when a new approach replaces traditional practice is difficult to define. For some time
it has not been good practice to use radium sources for intracavitary applications and it is now
not good practice to perform low dose rate (LDR) applications with Cs-137 sources.
The standard of care for patients with cervical carcinoma, as defined through Patterns of Care
studies, includes both external beam and intracavitary applications. 1 The historical dose
specification, based upon milligram-hours, is still used, although the actual dose prescription to
specific points is the norm. The sum of the external beam dose and the intracavitary dose is
routinely calculated to obtain a total dose, despite the fact that these are biologically very
different doses.
The physical properties of a Cs-137 source, namely a source diameter of 3.1 mm and a
physical length of 20 mm, limit the applicators available to the oncologist. Cs-137 applications
generally require the use of general anesthesia and 48–72 h of treatment, requiring the patient
to be confined to bed. The combination of general anesthesia and bed rest represents a risk to
the patient with a 6% life threatening complication rate and an overall mortality rate of 1.5%. 2
The appropriate number of LDR applications has never been established. The importance of
good geometry for an implant cannot be overstated.
High dose rate (HDR) intracavitary treatments are generally performed on an outpatient basis
using multiple fractions. The oncologist can specify multiple dose prescription points. The
dwell times and dwell positions can be optimized to meet this prescription. The dose
prescription can be modified for each fraction to customize the treatment. HDR applicators are
smaller in diameter, reflecting the smaller diameter HDR sources, and are designed to be
coupled together to provide a better geometry. The suggested number of fractions and the dose
per fraction have been published by the University of Wisconsin and others. 3
HDR intracavitary treatments have been successfully incorporated into routine treatment
management for patients with carcinoma of the cervix. They avoid the problems associated
with LDR treatments while offering the advantage of customizing the brachytherapy dose
distribution to the clinical situation. The smaller applicators increase the probability of a good
geometry and permit CT based dosimetric analyses. Multiple fractions offer the advantage of
brachytherapy treatments that can be modified as the tumor shrinks and the patient‘s anatomy
changes. The risks associated with a one or two fraction LDR procedure with prolonged bed
rest should no longer be tolerated.
Rebuttal
Dr. Palta expressed both general and specific concerns. His general concern about the lack of
randomized trials comparing HDR and LDR can be expanded to most developments in
radiation oncology. For example, there are no prospective, randomized trials comparing high
energy x rays with Co-60, or CT-based treatment planning with non-CT-based planning. This
concern is almost never addressed as treatment techniques evolve.
158
Dr. Palta‘s specific concern relates to ‗‗possible late radiobiological effects of using several
large doses of radiation.‘‘ Teshima et al. 4 reported complications in patients treated from
1975–1983 to be higher for the HDR group, although within acceptable levels, as compared
with the LDR group. Patel et al. 5 reported the results of their prospective, randomized HDR
versus LDR clinical trial, which was conducted from 1986–1989. Their only statistically
significant difference was a 19.9% incidence of overall rectal complications for the LDR arm
as compared with 6.4% for the HDR arm. These data support the American practice of more
fractions (5–12) and a lower dose per fraction. 6
Dr. Palta is correct in his observation that ‗‗HDR provides greater flexibility in optimizing the
dose distribution for each patient‘‘ for each application. We both agree that the importance of
the intracavitary application in managing patients with cervical carcinoma should never be
underestimated. In my opinion, the ability to customize the dose distribution for each
application is the strongest argument for HDR treatments. A fundamental belief in radiation
oncology is that better dosimetry results in better local control and lower complications.
The worldwide experience is substantial for HDR applications to treat patients with cervical
cancer. The North American experience continues to grow. It is my opinion that HDR
brachytherapy is clinically proven and, in the hands of those who have mastered its technical
challenges, offers significant dosimetric and patient care advantages that can no longer be
ignored.
REFERENCES
1. R. Lanciano et al., ‗‗Over 20 years of progress in radiation oncology: Cervical cancer,‘‘

Seminars in Radiation Oncology 7, 121–126 (1997).
2. K. Dusenbery et al., ‗‗Perioperative morbidity and mortality of gynecologic
brachytherapy,‘‘ Cancer (N.Y.) 67, 2786–2790 (1991).
3. A. Abitbol et al., ‗‗High dose rate brachytherapy for carcinoma of the cervix,‘‘ in High
Dose Rate Brachytherapy, edited by S. Nag (Futura, Armonk, NY, 1994), p. 355.
4. T. Teshima et al., ‗‗High-dose rate and low-dose rate intracavitary therapy for carcinoma
of the uterine cervix,‘‘ Cancer (N.Y.) 72, 2409–2414 (1993).
5. F. D. Patel et al., ‗‗Low dose rate vs high dose rate brachytherapy in the treatment of
carcinoma of the uterine cervix: A clinical trial,‘‘ Int. J. Radiat. Oncol., Biol., Phys. 28, 335–
341 (1993).
6. J. A. Stitt, ‗‗High dose rate brachytherapy for cervical carcinoma,‘‘ in Principles and
Practice of Brachytherapy, edited by S. Nag (Futura, Armonk, NY, 1997), pp. 505–513.
Against the proposition: Jatinder R. Palta
Opening Statement
Low dose rate brachytherapy has a long history of clinical use originating with preloaded
radium, to afterloading radium and cesium in the 1950s, and more recently to remote
afterloading of sources. Although the use of high dose rate (HDR) brachytherapy for cervical
cancer is increasing, low dose rate (LDR) brachytherapy is the most commonly used and most
extensively defined technique in the United States. It has been demonstrated unequivocally that
including LDR intracavitary brachytherapy in the treatment of cervical cancer improves the
159
survival rate for patients with advanced stage disease. Therefore, the importance of LDR
intracavitary therapy in the treatment of cervical cancer should never be underestimated.
The perceived advantages of HDR (improved dose distributions, reduced hospitalization of

patients, and decreased radiation hazards to personnel) are marred by a poor understanding of
the possible late radiobiological effects of using several large doses of radiation.
The classical radiobiological dilemma of large doses per fraction causing relatively severe late
damage to normal tissue has put the onus on proponents of HDR brachytherapy to demonstrate
that long-term complications with this modality are either less than or comparable to those for
LDR brachytherapy. Recent literature on this subject has focussed on finding the optimal
number of HDR brachytherapy fractions to limit the risk of late complications while achieving
local control rates equivalent to those for LDR techniques. HDR brachytherapy for cervical
cancer is still in its infancy. With a few exceptions, randomized trials have not compared HDR
and LDR brachytherapy. Most of the current controlled experience with HDR brachytherapy is
limited to a few institutions that have initiated in-house clinical protocols to compare HDR
results with historical LDR brachytherapy data for cervical cancer.
It is not prudent to extrapolate the experience of a few clinics to establish universal standards
of clinical practice. Randomized investigations are necessary to demonstrate that HDR
fractionation and dose schemes provide favorable rates of local control and decreased long-
term complications. HDR brachytherapy also requires a more rigorous set of dose distribution
criteria compared with LDR brachytherapy because HDR provides greater flexibility in
optimizing the dose distribution for each patient. At this time it is premature to consider LDR
brachytherapy with cesium sources, a proven treatment technique for cervical cancer for
decades, a breach of good practice, especially when the alternative HDR brachytherapy is still
unproven in the clinical arena.
Rebuttal
A new modality of treatment becomes a standard of good practice when its clinical efficacy,
cost benefit, safety, and preference by patients are well established. HDR brachytherapy for
intracavitary applications has yet to pass this test. Dr. Gillin cites a number of potential
advantages of HDR over LDR brachytherapy for cervical cancer, including the opportunity to
customize dose distribution, the outpatient nature of the procedure, decreased anesthesia
requirement, and greater patient convenience; but he fails to provide a convincing argument for
the advantage of HDR over LDR brachytherapy in terms of the most important criteria: tumor
control and risk of late normal-tissue complications. There is a paucity of randomized trials
comparing LDR and HDR brachytherapy. Most investigators who report results of HDR
therapy contend that they are comparable to those achieved with LDR therapy.
Dr. Gillin further argues that the physical properties of Cs-137 sources limit the applicators
available to radiation oncologists and the ability to customize dose distributions. This has not
been a problem over the past 30 years. Excellent results in terms of local control and survival
have established the benefits of conventional applicator systems for LDR treatment of cervical
cancer. The risk of perioperative mortality is negligible, less than 0.2% at most institutions.
The physical advantages of HDR over LDR can never compensate for the loss of the dose-rate
effect. It has been suggested that HDR treatment is more convenient to the patient. A recent
study1 has shown that patients prefer LDR brachytherapy because most women dislike pelvic
exams, especially when they involve manipulation of a tumorous uterus. Moreover, another
160
study2 has concluded that LDR brachytherapy is a more cost-efficient treatment in the United
States compared with other options, based on the RVUs for each option.
HDR intracavitary treatments have been incorporated into routine treatment management of
patients with cervical cancer only in Europe, Asia, and some third-world countries. The
impetus for HDR brachytherapy in those places is principally expeditious treatment. It is
premature to make a drastic change in the treatment approach in the United States, in light of
the large degree of uncertainty about the dose and fractionation of HDR brachytherapy
required to provide control tumor without major late complications.
REFERENCES
1. J. Wright, G. Jones, T. Whelan, and H. Lukka, ‗‗Patient preference for high or low dose
rate brachytherapy in carcinoma of the cervix,‘‘ Radiother. Oncol. 33, 187–194 (1994).
2. A. A. Konski, P. M. Bracy, S. G. Jurs, S. J. Weiss, and S. R. Zeidner, ‗‗Cost
minimization analysis of various treatment options for surgical stage I endometrial
carcinoma,‘‘ Int. J. Radiat. Oncol., Biol., Phys. 37, 367–373 (1997).
161
3.2. Reimbursement for high dose rate brachytherapy should be

based on the number of dwell positions of the source during
treatment
Geoffrey S. Ibbott and James M. Hevezi

Reproduced from Medical Physics, Vol. 26, Issue 3, pp. 347-349 , March 1999
OVERVIEW
The charge for a radioactive seed implant for cancer treatment is based in part on the number
of seeds used in the implant. This procedure is justifiable because there is an acquisition cost
per seed, and the complexity of implanting the seeds and estimating the resulting dose
distribution increases with the number of seeds. In many institutions, this charge model has
been followed in high dose rate (HDR) brachytherapy where reimbursement is based on the
number of ‗‗dwell‘‘ positions of a single source employed for treatment. For example, a source
moved in 0.5 cm increments yields a higher charge than one moved in 1 cm increments, even
though the finer increments may not yield a significant improvement in dose distribution. This
Point/Counterpoint article explores the legitimacy of this practice.
Arguing for the Proposition is Geoff Ibbott. Dr. Ibbott began his career in medical physics at
the University of Colorado Medical Center in Denver in 1968. He received an M.S. degree in
medical physics from the same institution in 1981 and received a Ph.D. in Radiation Biology
from Colorado State University in 1993. He has been Director of Medical Physics at the
University of Kentucky since 1994 and is the President of the American Association of
Physicists in Medicine. Dr. Ibbott has served on several committees of the AAPM Professional
Council and was chair of the Council from 1993 to 1997, during which time he developed an
interest in reimbursement issues.
Arguing against the proposition is James M. Hevezi. Dr. Hevezi spent the early part of his
career in Diagnostic Radiological Physics at M. D. Anderson Hospital (1970– 1979), and
switched to Radiation Therapy Physics at the University of Arizona in the early 1980s. There
he worked in clinical hyperthermia methods to treat cancer. In 1983 he moved to the Phoenix
metro area as vice president of Radiological Physics Services. There he helped develop the first
Nucletron HDR source procedure to treat bronchogenic cancer and other applications. He is
currently Director of Medial Physics at the Cancer Therapy & Research Center in San Antonio
and holds a faculty position at the University of Texas Health Science Center at San Antonio.
Dr. Hevezi serves as Chair of the Economics Committee of the ACR Commission on Medical
Physics and has been active in determining policy for medical physicists in this area.
For the proposition: Geoffrey S. Ibbott
Opening Statement
Reimbursement for radiation oncology is classified under codes published in the CPT manual. 1
Providers and payers of medical services including HCFA, the agency that administers the
Medicare program, use this guide almost universally. Other than the CPT manual itself, there is
162
a paucity of published assistance regarding use of CPT codes for HDR brachytherapy. The
ACR published a User‘s Guide a number of years ago,2 but it now is out of date, and the HDR
CPT codes were not published until after the ACR guide appeared.
Four CPT codes are used for HDR, where the number of ‗‗source positions or catheters‘‘
determines the complexity. The descriptions indicate that reimbursement is tied to the number
of dwell positions. Reimbursement for brachytherapy treatment planning is requested using
CPT codes 77326 to 77328 where, for HDR, the manual indicates that the complexity of the
plan is tied to the number of ‗‗sources.‘‘ This terminology should be interpreted as referring to
HDR dwell positions, as HDR units have only one source.
Dose planning includes the review of localization films and performance of calculations to
determine the number of dwell positions, their locations, and the dwell time at each position. In
many departments, the use of 1 cm spacing between dwell positions is customary. A large
target volume might require many dwell positions, perhaps distributed among several catheters.
The more complex the arrangement, the more effort is required to determine the ideal source
configuration and to verify the accuracy of the calculations.
The dwell position information is transferred to the console of the HDR unit and the accuracy
of the transfer is verified. Prior to treatment delivery, the position of the applicator or catheters
may be verified by fluoroscopy. A more complex arrangement requires more effort to confirm
to the physician‘s satisfaction that the position of the applicator or catheters is correct. When
delivery begins, the staff must monitor the patient and the position of the source. Treatments
involving more dwell positions generally are more complex and take more time. Consequently,
the cost to the facility increases with number of dwell positions. The time and effort required
by the technical staff and the physicist are increased, and to a small extent, supplies such as
catheters also increase. The effort exerted by the physician also increases as the complexity of
the procedure increases.
As was suggested in the Overview, the linking of reimbursement to the number of dwell
positions may bring about the temptation to artificially increase the dwell positions over the
number required for an acceptable dose distribution. Such practice is inappropriate and is to be
discouraged, just as the unnecessary use of complex external beam therapy should be avoided
when a simpler field arrangement yields a satisfactory dose distribution.
Rebuttal
Dr. Hevezi‘s recommendation is to base reimbursement on the number of catheters used for an
HDR treatment. The tissue volume treated with HDR units are three dimensional, and in many
cases, the number of catheters used determines two of the dimensions. But the number of dwell
positions determines the third dimension. Generally, the total number of dwell positions
increases with the size of the target volume. Basing reimbursement on the number of catheters
would roughly maintain the proportionality between complexity and number of dwell
positions, but would reduce reimbursement overall.
Dr. Hevezi refers to a new Radiation Oncology Draft Manual for Policy Development. This
document, as the title indicates, is still in draft form and is not yet policy so is not presently
available to guide reimbursement. However, as he points out, one provision would allow
uncoupling of the level of complexity of the technical component from that of the professional
component. This would be an appropriate change, but unfortunately it would be of little benefit
163
to hospitals and other entities that bill the technical component separately. Today at least, the
technical relative values (RVUs) are the same for all four reimbursement codes for HDR.
Technical reimbursement from Medicare is independent of the level of complexity!
Professional reimbursement does vary with complexity, but the RVUs are quite small, and
there would be little to gain by escalating the complexity level.
Rather than changing the reimbursement policies arbitrarily, a complete analysis should be
done of the cost of the procedure. This can be done, as has been done for the procedures and
charge codes, by surveying a representative sample of treatment centers. The cost of the
equipment, supplies, and attendant personnel (some mandated by NRC Regulations) must be
considered. The relationship between reimbursement and complexity level (based either on
number of catheters or number of dwell positions) then can be determined in a realistic and
unambiguous fashion.
REFERENCES
1. Current Procedural Terminology (CPT ‘99), 4th ed., American Medical Association,
Chicago, IL, 1998.
2. User‘s Guide from the Radiation Therapy Relative Value Scale, American College of
Radiology, Reston, VA, 1989.
Against the proposition: James M. Hevezi
Opening Statement
Low dose rate (LDR) interstitial or intracavitary brachytherapy is a common procedure.

Recently, the CPT manual1 has indicated that the complexity of the procedure should be tied to
the number of sources involved in the procedure. As High Dose Rate (HDR) procedures have
become prevalent, similar criteria for complexity have been entered into the CPT manual.
Instead of the number of sources, however, the complexity of the procedure is tied to the
number of dwell positions of a single source. Although this discussion specifically addresses
HDR procedures, policy decisions for HDR reimbursement should not be separated from those
for LDR procedures.
There are two levels of reimbursement guided by whether the procedures are hospital-based,
in-patient procedures or, either hospital-based, out-patient or freestanding centerbased out-
patient procedures. This discussion is confined to the latter two, i.e, fee-for-service CPT-
directed reimbursement (versus DRG based reimbursement). Most LDR brachytherapy
procedures involve introduction of many sources to achieve an appropriate dose rate and
distribution to the target. Usually, the number of sources exceeds the limit required for
definition as a complex planning and delivery scheme. Exceptions are simple, gynecological
procedures using Cesium sources, but these are generally DRG driven in-patient procedures.
Most out-patient brachytherapy procedures should be scaled to the number of needles or
catheters used. In this manner, implants that require only a few catheters or needles, whether
HDR or LDR, will be allotted to the simple or intermediate realm, and those that require many
needles or catheters will be relegated to the more complex reimbursement levels.
164
In the new Radiation Oncology Draft Manual for Policy Development, 2 several of the
professional societies support consigning the level of procedure complexity to the number of
needle/catheter tracks rather than to the number of individual sources used in the implant. With
the new paradigm, procedures such as ultrasonically guided prostate seed implants will not
change their level of complexity because they use enough needle tracks for the procedure to be
classified as complex.
The situation is more acutely impacted for HDR procedures where single or double catheter
tracks would be relegated to reimbursement as a simple procedure. HDR procedures involving
the use of multiple catheters, such as template techniques or the HDR version of prostate
implants, would still be complex. A point to emphasize is the ‗‗uncoupling‘‘ of procedure
segments. Prior to the recommendations of the new Radiation Oncology Draft Manual for
Policy Development, if one segment of a procedure was complex the entire procedure was
classified as complex. Now, each segment is to be evaluated as its own complexity level for
reimbursement considerations. Also, the professional segment could be evaluated as simple or
intermediate, while the technical effort could be complex. Or the reverse could be the case. The
question then becomes, does it take more effort to produce a plan/treatment for a complex
HDR procedure compared to an intermediate or simple procedure? There will be financial
separation between these entities in a catheter/ needle based approach versus a dwell
position/source based approach. It is clear that a single catheter in an HDR bronchial
application, for example, requires less effort than a 20 catheter gynecological or prostate
template procedure. The catheter/needle approach works for both HDR and LDR procedures.
Rebuttal
Dr. Ibbott does not address the key issue here. Does the use of dwell positions to drive
procedure complexity actually result in a separation of appropriate reimbursement levels for
HDR procedures? And, if it does not, can we identify another reimbursement paradigm that
will effect this separation and, at the same time, fairly recompense each segment of the
procedure? Clearly the use of the needle/ribbon paradigm of driving procedure complexity for
reimbursement purposes will result in a better separation of these entities—both for HDR and
LDR brachytherapy procedures. In addition, using the needle/ribbon paradigm for
reimbursement levels will discourage the practice of utilizing 0.5 cm or 0.25 cm dwell position
separations in order to drive the procedure into a complex level for reimbursement purposes.
This practice was alluded to by Dr. Ibbott at the end of his opening statement and, although he
appropriately cautioned individual users against artificially increasing the number of dwell
positions when unnecessary, his caution will not result in reducing these practices. Linking
procedure complexity to the number of needle/ribbon positions will allow those individuals
who wish to use smaller dwell position separations to do so without fear of reprisal while
satisfying their clinical decision making process. Many HDR planning systems allow the user
to merely fill all of the dwell positions available for catheter tracks with a single selection
keystroke. This is an efficient way to automatically load HDR catheter tracks, but generally
results in the smaller separation and many dwell positions.
REFERENCES
1. Current Procedural Terminology (CPT ‘99), 4th ed., American Medical Association,
Chicago, IL, 1998.
2. ACR unpublished draft (AAPM, ACMP, ASTRO, ACRO, AFROC text), 1998.
165
3.3. Calculation of brachytherapy doses does not need TG-43

factorization
Haijun Song and Gary Luxton

OVERVIEW
In brachytherapy, data needed to derive the radial dose function (RDF) and anisotropy function
(AF) (as defined in AAPM TG-43) are equivalent to data in the relative along-away table
(RAAT). In fact, the RAAT data can be derived from the RDF and AF, if a geometry factor is
assumed. The TG-43 approach involves factorization and reproduction of the RAAT. An
alternate approach is to use the RAAT directly along with the dose rate constant (DRC) in TG-
43. Both approaches yield the same dose calculation. Since RAAT can be easily obtained (e.g.,
with radiochromic film), the direct RAAT approach is simpler and the TG-43 factorization is
unnecessarily complicated. We propose using RAAT directly instead of the TG-43
factorization.
Arguing for the Proposition is Haijun Song, Ph.D. Dr. Song developed a miniature neutron
probe [Med. Phys. 29, 15–25 (2002)] for his Ph.D. thesis at MIT. He held postdoctoral
positions at the Laboratory for Accelerator Beam Applications at MIT and the Center for
Radiological Research at Columbia University, before he trained in medical physics at the
Thomas Jefferson University Hospital. He was part of a team that performed more than one
hundred prostate brachytherapy implants in one year. He held a faculty position at the Jefferson
Medical College of Thomas Jefferson University, before recently joining the faculty of the
School of Medicine of Yale University.
Arguing against the Proposition is Gary Luxton, Ph.D. Dr. Luxton received his Ph.D. in
physics from Caltech in 1970, then did research in experimental particle physics at SLAC,
Argonne and Fermilab. Beginning in 1975, he worked at Stanford on a design for negative pi-
meson radiation therapy. He is certified by the ABR. As director of physics at USC/Norris
Comprehensive Cancer Center, he led development of programs in ophthalmic plaque
brachytherapy and radiosurgery. He returned to Stanford in 1998 as Associate Professor, where
he is presently Chief of Clinical Physics. Current interests include radiobiological modeling of
IMRT, and quality assurance for IMRT and robotic radiation delivery.
FOR THE PROPOSITION: Haijun Song, Ph.D.
Opening Statement
The Dose Rate Table (DRT) for a brachytherapy source can be used as a look-up table to
calculate dose distributions in the clinic. The DRT is the same as the traditional along-away-
table, one example of which is for the Cf-252 needle shown in Table 1 of Ref. 1.
166
The AAPM TG-43 protocol takes a "dissemble-and-assemble" factorization approach: The

Dose Rate Constant (DRC) and other factors, including the Radial Dose Function (RDF), the
Anisotropy Function (AF) and the Geometry Factor (GF), are derived from the measured DRT.
In clinical practice, these factors are multiplied to reproduce the measured DRT. While the
DRC improves the accuracy of dose determinations over the pre-TG-43 exposure rate constant,
the advantage of adopting the other factors is debatable.
Since the GF tends to dominate the dosimetric properties in the immediate vicinity of a
brachytherapy source, the RDF, with the GF accounted for, offers a better tool than the DRT
for assessing radiation quality. However, such an assessment is needed only during the early
phase of selecting a particular isotope for clinical use. One example might be assessment of the
numerous brands of I-125 seeds, for which the TG-43 factors are usually published without
comparison with other brands or other isotope species.
Under the point isotropic source approximation, the AF is reduced from a two-dimensional
matrix to a one-dimensional anisotropy factor, or even further reduced to a single anisotropy
constant. The TG-43 dose calculation formula takes the simple form of 1/r2. However this
approximation is accurate only for energetic radiation at large distances. This situation does not
call for a protocol with the sophistication of the TG-43. For low energy brachytherapy sources,
the anisotropy constant will be phased out.2 Also note that the angular dependence can be
averaged out for the DRT in the same way that the AF is reduced to the anisotropy factor.
A full-blown two-dimensional AF has the same grid points as the measured DRT. The same
look-up and interpolation will be performed on the AF and the DRT. Extra calculations are
needed for the GF and the RDF. Thus dose calculations using the DRT are more efficient.
The GF is based on assumed geometry, which is sometimes an approximation of the "true"

geometry for convenience. A "wrong" geometry does not affect the dose accuracy so long as
identical GFs are used by the factor-generating people and the dose-calculating people. This
shows the arbitrariness and dosimetric irrelevance of the GF. In reality the GF is not always
published along with the other factors, and error can be introduced when the seed geometry is
approximated differently.
Because of the GF arbitrariness and the inter-dependence of the RDF, AF and GF, uncertainties
of the RDF and AF cannot be determined based on the uncertainty of the DRT. With
uncertainties unknown, the value of the RDF and the AF as a "gauge" for radiation quality and
characteristics of seed construction is compromised.
In summary, dose calculations for brachytherapy sources can be done more efficiently by using
the DRT directly, or equivalently a normalized (to the reference point) DRT with the DRC.
Where it is valid, the point isotropic source approximation offers a convenient dose calculation
formula. However the TG-43 would be an over-kill if its sole benefit were to provide the point
isotropic source approximation.
Rebuttal
The TG-43 factorization is not necessary in the sense that 2-D dosimetry, which is required for
cylindrical sources as in Ref. 1, can be performed without it. With regard to relying on intuition
for interpretation of dosimetry data, thereby contributing to QA, manual calculations, etc., this
benefit does not seem to be readily accessible. First, the radial dose and anisotropy functions
167
for 125I and 103Pd seeds are steep curves and, the anisotropy for high-energy 192Ir seeds can
deviate 20% from unity. Intuitive right or wrong judgments cannot be made for these curves.
Variations in the radial dose and anisotropy functions make manual calculations cumbersome,
if possible at all. Second, aside from the potential problems stated in the Position Statement, the
geometry factor can be a challenge for manual calculations, even with the help of computers.
For example, try an evenly distributed cylinder that is 3 mm long. Then try stacking 14 of these
cylinders in a string with 1 mm separations. Then try even more complicated activity
distributions.
The geometry factor does not account for differences such as construction materials, x-ray
production and attenuation. If we go down the road of factorization, we will have to introduce a
"construction factor" to remove brand dependence from the radial dose and anisotropy
functions for different brands of seeds made of the same isotope. And that would introduce
another unnecessary burden for dose calculations.
Compliance with the TG-43 factorization by major brachytherapy planning systems and the
physics community should not be taken as evidence for or against the necessity of the TG-43
factorization.
AGAINST THE PROPOSITION: Gary Luxton, Ph.D.
Opening Statement
Brachytherapy dose calculations are steeped in tradition, and when TG-433 was convened, the
majority of clinical 125I and 192Ir seed implants were being calculated with point source dose
models. One issue addressed by the Task Group was to recommend how clinical brachytherapy
dose calculations should take into account the 2-D properties of cylindrically-symmetric seeds
just then being measured or calculated using Monte Carlo, and how to implement the
recommendation. This was not an entirely trivial task. It was widely recognized that accuracy
could be significantly improved, particularly for small-volume treatments such as radioactive
plaque therapy for ocular melanoma or interstitial brain implants. Commercial brachytherapy
planning systems, however, offered little support for 2-D seed dose models. The problem was
one of providing a smooth transition path from 1-D to 2-D seed source models that would
actually be used clinically and supported by commercial brachytherapy planning software.
A solution was adopted in the form of the factorization algorithm, representing the 2-D dose
distribution from a seed as a product of three factors. One factor extracted the geometric
modeling of the distribution of source material including the inverse-square law, another
accounted for attenuation and scatter build-up behavior as a function of transverse distance
from the source, while the third, the anisotropy function, depended on both angle and distance.
By extracting the line source or other geometry factor, the remaining factors readily lent
themselves to physical interpretation, which was worthwhile from the point of view of error
detection and quality assurance in both commissioning and daily practice.
Dose rate behavior in close proximity of a source can be understood by the geometry factor.
The slowly-varying behavior of the anisotropy function can be approximated by a few values,
enabling the formalism to provide convenient manual calculation checking capability. Physical
interpretation of a source model lends itself to intuitive and rapid source data checking, with
only limited data needed to achieve reasonably accurate dose calculation. The factorization
168
model also offers seamless attachment to traditional point source model calculations through
the anisotropy constant.
Certainly, the model could be replaced by a rectangular grid of relative dose rates. Bare
numbers are not particularly intuitive, however. Verification of data would be less transparent
and the medical physicist's ability to design an implant would not be facilitated.
The TG-43 factorization model appears to be supported by all major brachytherapy planning
systems, and much seed data has now appeared in this form, such as in Refs. 4 and 5. The
model has robustly been applied to the emergent technology of gamma source intravascular
brachytherapy.6 To illustrate its usefulness, we note that for five different source models, in
close proximity (2 mm) to an HDR 192Ir source, dose rate per unit source strength is
approximated to within 2–3% over a large fraction of total solid angle by the geometry factor
multiplied by the point source dose rate constant.7
Rebuttal
Dr. Song suggests that TG-43 factorization for brachytherapy source dosimetry is deficient
because it requires extra interpolation calculations, is unnecessarily sophisticated and subject to
operator error, and is needed only during early clinical isotope evaluation. We consider these
objections in turn.
Using several factors in place of a single 2-D function does require more interpolations in
computer calculations. However, this is a minor cost for the added clarity of a geometric model.
Computer interpolations are fast, and most of the cpu time is spent on calculating seed and
calculation-point orientation vectors, which are needed in the DRT method as well. The
offsetting advantage in TG-43 is its greater ease in performing accurate manual calculations.
The argument is made that TG-43 is unnecessarily sophisticated for large r, for which the
geometric factor reduces to 1/r2. There remains a need here, however, for the RDF, g(r), which
is more easily visualized and understood than is a Cartesian matrix. Dr. Song's reference to a
plan to phase out the anisotropy constant for the 1-D model actually refers to retaining
accuracy by using the distance-dependent anisotropy factor.
TG-43 algorithm calculations are stated to be subject to possible errors because a different
geometric factor may be used to construct dose tables than is used for dose calculations. The
line-source model is a good approximation for practical sources; only in very demanding
applications might a different model be used for which a significant difference could arise.
Given such a circumstance, it is unlikely for an error of the type described to be made. In
producing and using a source model, error is not less likely in transcribing and interpolating
DRT matrices.
Complete TG-43 sophistication may indeed be necessary only during clinical isotope
evaluation, but this does not detract from its making available an advanced, easily-calculated
source model. No model is perfect, but TG-43 factorization has proven to be resilient, and has
served the brachytherapy community well.
REFERENCES
169
1. R. D. Colvett, H. H. Rossi, and V. Krishnaswamy, "Dose distribution around a Cf-252

Needle," Phys. Med. Biol. 17, 356–364 (1972).
2. R. Nath, M. Rivard, B. Coursey, L. DeWard, W. Hanson, M. Huq, G. Ibbott, and J.
Williamson, "Status of the AAPM radiation therapy committee's subcommittee on low-energy
interstitial brachytherapy source dosimetry: Procedure for the development of consensus
single-source dose distributions," Med. Phys. 29, 1349 (2002).
3. R. Nath, L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S.
Meigooni, "Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM
Radiation Therapy Committee Task Group No. 43," Med. Phys. 22, 209–234 (1995).
4. A. S. Meigooni, K. Sowards, and M. Saldano, "Dosimetric characteristics of the
InterSource 103palladium brachytherapy source," Med. Phys. 27, 1093–1100 (2000). R. S.
Sloboda and G. V. Menon, "Experimental determination of the anisotropy function and
anisotropy factor for model 6711 I-125 seeds," Med. Phys. 27, 1789–1799 (2000).
5. R. Nath, H. Amols, C. Coffey, D. Duggan, S. Jani, Z. Li, M. Schell, C. Soares, J.
Whiting, P. E. Cole, I. Crocker, and R. Schwartz, "Intravascular brachytherapy physics: Report
of the AAPM Radiation Therapy Committee Task Group No. 60," Med. Phys. 26, 119–152
(1999).
6. P. Papagiannis, A. Angelopoulos, E. Pantelis, L. Sakelliou, D. Baltas, P. Karaiskos, P.
Sandilos, and L. Vlachos, "Dosimetry comparison of 192Ir sources," Med. Phys. 29, 2239–2246
(2002).
170
CHAPTER 4
General Imaging: Image Quality, Magnetic

Resonance, Ultrasound
4.1. The development of technologies for molecular imaging should
be driven principally by biological questions to be addressed rather
than by scaling down macro-level imaging technologies
Gary D. Fullerton and John D. Hazle

OVERVIEW
A recent review of Molecular Imaging1 suggests that fundamental biology, rather than design
engineering, should guide the development of imaging technologies at the molecular level.
However, the history of biomedical imaging reveals a pattern of exploiting technology
developments in other fields and adapting them into biomedical research and clinical care. These
two different approaches to technology development are the subject of this month's
Point/Counterpoint.
Arguing for the Proposition is Gary Fullerton, Ph.D. Dr. Fullerton, Ph.D., a graduate of the
University of Wisconsin medical physics program, serves as the Malcolm Jones Distinguished
Professor of Radiology at the University of Texas Health Science Center at San Antonio. As
founding director of the Radiological Sciences graduate program, he directs 60 faculty members
in supervising the studies of 52 students. Dr. Fullerton was secretary of AAPM in 1981–1983,
and president in 1991. He has served as president of the Society of Magnetic Resonance Imaging
and founding Editor-in-Chief of the Journal of Magnetic Resonance Imaging 1990–2000. He was
co-president of the World Congress on Medical Physics and Bioengineering in 1988 and served
as Secretary General of the IOMP and IUPESM 1997–2003. Dr. Fullerton's present research
focus is on translation of biomolecular imaging into clinical practice.
Arguing against the Proposition is John Hazle, Ph.D. Dr. Hazle received the Ph.D. from The
University of Texas Graduate School of Biomedical Sciences. He joined the M. D. Anderson
faculty in 1989 as Assistant Professor in the Department of Radiation Physics. He is now
Professor and the first Chair of the Department of Imaging Physics. John is certified in diagnostic
radiological, therapeutic and magnetic resonance physics. His research interests are minimally-
invasive image-guided therapies and small animal imaging. He is Director of the NCI-funded
171
Small Animal Cancer Imaging Research Facility and a driving force in the creation of a new
Center for Advanced Biomedical Imaging Research at M. D. Anderson.
FOR THE PROPOSITION: Gary Fullerton, Ph.D.
Opening Statement
The statement, "The development of technologies for Molecular Imaging should be driven
principally by biological questions to be addressed rather than by simply modifying existing
imaging technologies," proposes a fundamental shift in the paradigm responsible for the
development of medical imaging since the discovery of x rays. Over the 20th century, medical
imaging development was driven by technological innovation and engineering improvements in
physical equipment. A new imaging modality driven by development of the biological
knowledge base represents a fundamental change. Understanding such an important change is of
utmost importance to medical physicists working in imaging research. The measure of success,
and the economic growth of medical imaging, reside firmly in the ability to implement new
procedures with higher diagnostic specificity and sensitivity. Such imaging methods, though
expensive in themselves, can in many cases provide cost-competitive resolution of patient
healthcare problems.
We will consider three arguments; (1) the opinion of leaders in molecular imaging, (2) the
scientific focus of self-proclaimed molecular imaging investigators and (3) the major NIH
investment in genomic and proteomics research. These arguments lead me to conclude that the
statement by Dr. Piwnica-Worms is correct. Medical physicists must either learn to include the
biology of molecular imaging in their research programs or prepare to become irrelevant to the
future of radiology.
A search of the molecular imaging literature leads to a small circle of key individuals and
institutions. They are the leaders of the Society of Molecular Imaging which has the stated
purpose, "An international scientific educational organization whose purpose is to advance our
understanding of biology and medicine through noninvasive in vivo investigation of cellular
molecular events involved in normal and pathologic processes."2,3 It is clear from this statement
and from the web pages of the leaders of molecular imaging that they as a group believe that
biology rather than an underlying technology is driving their field.
A review of the most recent program of the Society of Molecular Imaging4 shows a primary
focus on biological questions, design of optical or radioactive molecular imaging probes and/or
potential clinical applications. Less than 10% of the program (only one of twelve major
symposia) is devoted to technological development, while the remaining 90% is devoted to
biology and chemistry. The ACR Primer on Molecular Imaging,5 a commercial special issue on
Molecular Imaging,6 and the more popular radiology news journals7 all share this focus on
biology. The bulk of specialists in molecular imaging are conducting studies of biology using
imaging as the tool of choice.
The final consideration is the substantial refocus of federal funding on genomics, proteomics and
the relation of these concepts to medicine and health care. Examples of the growing significance
of these areas to world science are the special issue focus of journals such as Science8 and
Nature9 on the fundamental importance of these discoveries.
172
All three arguments support the contention that biological knowledge, rather than technological
evolution of existing devices, dominates the development of molecular imaging. This is a
recognizable shift from the development of imaging in the 20th century. It holds great promise
for increasing the importance of medical imaging in healthcare decision making around the
world.
Rebuttal
I agree with many of the points made by my colleague, but he has missed the main point of the
discussion, which is "should technology development be driven by biological questions or the
evolution of existing imaging technologies?" There are two examples that clarify the magnitude
of this omission. First consider the only session on technology at the 2004 Society of Molecular
Imaging Congress which had the following objective: "To address molecular and small animal
imaging challenges, revolutionary developments have occurred in hardware and software. This
session is designed to highlight the newest cutting edge developments in instrumentation and
new strategies that will enable us to "see" more with greater sensitivity and specificity."
Existing imaging technologies are being reconfigured and adapted to the size and biology of
mouse and rat models. Rodent biology is driving technology development.
A second example is the introduction of human CT/PET units for clinical molecular imaging.
These instruments are optimized to integrate anatomical information from CT with molecular
functional information from PET. The medical need for co-registration of anatomical and
functional information has driven the development of the technology. I readily accept that
molecular imaging cannot prosper without continued improvement of imaging technology. The
direction of these developments, however, will be dictated by biological questions that can only
be answered through molecular imaging.
AGAINST THE PROPOSITION: John Hazle, Ph.D.
Opening Statement
At the end of the first Michigan State workshop on Molecular Imaging in 2001, Dr. Elias
Zerhouni moderated a discussion to define molecular imaging in the context of creating a new
scientific society (Society for Molecular Imaging). The discussion was lively, and eventually a
vision for the new society was adopted consistent with the definition of molecular imaging
accepted by The American College of Radiology's Commission on Molecular Imaging. The ACR
definition is: "Molecular imaging may be defined as the spatially localized and/or temporally
resolved sensing of molecular and cellular processes in vivo."
I will argue that for most in vivo applications the best approach to developing molecular imaging
hardware is to optimize existing technologies. It is relatively clear that new molecular probes
have a significant probability of yielding leap-step advances. The lofty goal of molecular
imaging cannot be achieved without significant advances in instrumentation as well.
The two imaging modalities that will likely have the most impact in patients over the next decade
are positron emission tomography (PET) and magnetic resonance (MR). For PET, sensitivity has
been an issue since the technology's inception. Detector composition, electronic component
capability and instrument geometries have all been improved in order to enhance sensitivity.
173
Advances in crystal design and composition still hold promise for improving overall sensitivity.
New experimental systems and flat-panel detectors may provide significant improvements in
sensitivity. Further, advances in electronics are permitting new counting strategies, and advances
in computing power are allowing for new reconstruction algorithms with enhanced sensitivity.10
Biology will drive the development of new probes that will create opportunities for rapid
advancement. Refining existing technologies to detect positron-induced annihilation radiation is
the approach most likely to yield significant results for instrumentation.
The case is much the same for MR. Advances in magnet design are allowing higher field
strengths for both animal and human imaging. There are now several whole-body instruments
operating at 7 T, and there is an 8 T system at Ohio State and a 9.4 T system at the University
of Illinois at Chicago. These magnets provide improved sensitivity, with detectability of low-
contrast detail improving by a factor of 10–15 in going from 1.5 T to 9.4 T. Coupled with
advances in coil and electronics performance, a 20–30 fold increase in sensitivity may be
achievable. The real opportunity for leap-step improvement in detectability is the development of
field-strength-specific contrast agents.11 Although the potential sensitivity of new probes may be
10–100 fold, the mechanism of detection is likely to be a permutation of an existing approach
(pulse sequence of acquisition strategy). That is, continued development of more mature
instrumentation is likely to yield the most significant advances in the modality.
Optical imaging is an exception in this debate. Here probes and detector systems are tightly
coupled. The potential is present to find a new probe system based on biology that requires a new
detector technology. This potential could drive instrumentation development. Nevertheless, there
is still a lot of potential in continuing to optimize existing technologies for improved sensitivity
and depth resolution.
In conclusion, I believe the development of the instrumentation used for molecular imaging is
best driven by continued optimization of existing instrumentation for known tracers and contrast
mechanisms.
Rebuttal
I whole-heartedly agree that medical physicists must learn more biology, including physiology
and metabolism, to pursue research in molecular imaging. However, we need to remember that
our core value comes from the deep technical knowledge we have of the instrumentation (i.e.,
tools) used to detect these processes, and the optimization of the performance of these devices.
To fully exploit this knowledge we must understand the biological questions to a degree greater
than most of us were prepared for in our educational careers. Second, the focus on genomics and
proteomics is not surprising because these are new fields with broad potential applications. The
molecular imaging community needs to develop ways to exploit advances in these areas with
new probes and tracers. Unless these systems require fundamentally new imaging techniques, we
will need to incrementally build better instrumentation technologies for molecular imaging.
REFERENCES
1. D.R. Piwnica-Worms, "Introduction to Molecular Imaging," Supplement to J. Am. Coll.

Radiol. 1(1S), 2–3 (2004).
2. Ralph Weissleder, Center for Molecular Imaging, http://www.mgh-cmir.org/.
3. Michael Phelps, Crump Institute for Molecular Imaging,
http://www.crump.ucla.edu/index.tcl.
174
4. Society of Molecular Imaging, 2004 Scientific Program,

http://www.molecularimaging.org/pdf/SMI2004finalprogram.pdf.
5. J.H. Thrall, "ACR Primer on Molecular Imaging," JACR 1(1S), 1–32 (2004).
6. Philips, "Molecular Imaging: the road ahead," Med. Mundi 47/1, 1–88 (2003).
7. J. Brice, "Molecular imaging transports diagnosis to the next level," Diagn. Imaging
(2001), http://www.diagnosticimaging.com/specialedition/molecular.shtml.
8. Biological Imaging, Science 300, 1–196 (2003).
9. Human Genomics and Medicine, Nature (London) 496, 439–481 (2004).
10. W. H. Wong et al., "Feasibility of a high-speed gamma-camera design using the high-
yield-pileup-event-recovery method," J. Nucl. Med. 42, 624–632 (2001).
11. D. M. Doble et al., "Toward optimized high-relaxivity MRI agents: the effect of ligand
basicity on the thermodynamic stability of hexadentate hydroxypyridonate/catecholate
gadolinium(III) complexes," Inorg. Chem. 42, 4930–4937 (2003).
175
4.2. Functional MRI is fundamentally limited by an inadequate

understanding of the origin of fMRI signals in tissue
John C. Gore and Robert W. Prost

OVERVIEW
Functional magnetic resonance imaging (fMRI) is an exciting and dynamic area of imaging
research. Many proponents of this technology believe that its clinical applications will yield
major breakthroughs in the early diagnosis of brain abnormalities, mental conditions and
behavioral deficiencies. Others think that these major breakthroughs will be severely limited by
the phenomenological nature of fMRI research, and progress will be curtailed until the
physiological origin of fMRI signals is understood. This difference of opinion is explored in this
Arguing for the Proposition is John C. Gore, Ph.D. Dr. Gore is Chancellor's University Professor
of Radiology and Biomedical Engineering, as well as Physics and Molecular Physiology and
Biophysics, at Vanderbilt University. He trained in Physics in London, and has worked in
medical imaging research for over 30 years, and in the field of MRI since 1979. His research
interests include the use of NMR to study tissue biophysics and factors that modulate MRI
signals from tissues, such as the BOLD effect. He has contributed over 300 papers and chapters
to the imaging literature, including many studies using fMRI to investigate human cognition.
After 20 years at Yale University, he moved in 2002 to direct a new Institute of Imaging Science
at Vanderbilt University.
Arguing against the Proposition is Robert W. Prost, Ph.D. Dr. Prost is an Assistant Professor of
Radiology at the Medical College of Wisconsin, where he serves as Chief of MR Technical
Advances. In that capacity he is responsible for technical aspects of translational MR research,
including pulse sequences, post-processing and RF coil development. Before joining the Medical
College in 1992, he was Lead MR System Designer for General Electric Medical Systems. Dr.
Prost's primary research concerns are in functional MR imaging and low-field MR spectroscopy.
FOR THE PROPOSITION: John C. Gore, Ph.D.
Opening Statement
FMRI indirectly detects blood flow and oxygenation changes that accompany neural activity. For
some applications such as neurosurgical planning, the primary aim of fMRI is to accurately
depict the spatial locations of regions selectively recruited after a stimulus. In these cases, a
precise interpretation of the blood oxygenated level dependent (BOLD) signal is not essential. In
other contexts, however, such an empirical approach is fraught with potential dangers of
misinterpretation. It is also likely to overlook valuable additional information that may be
obtainable. In studies of human cognition, for example, maps may be recorded depicting
176
distributed networks of regions that respond to a specific task. Inferences may then be drawn
about the significance of the relative weights of different nodes within the network, and about
differences in the amplitude and area of activation among groups or conditions. Implicit is the
premise that the BOLD signal reflects, at least semi-quantitatively, the degree of neural activity.
However, we currently have little basis on which to interpret the amplitude of BOLD signals.
Numerous studies show that the BOLD signal changes with behavior, learning, intervention and
physiology, but does not respond linearly to stimulus intensity.1 The relationship between
electrical discharges and BOLD signal is confounded by unknowns in essential couplings at
different stages (e.g., between electrical and synaptic activity of neurotransmission, between
metabolic demand and blood flow and oxygen use, and between hemodynamic changes, vascular
architecture and the NMR signal). The BOLD signal is affected by several factors and
physiological variables that are not yet understood. It is well known, for example, that baseline
vasodilation (e.g., breathing air enriched with carbon dioxide, or even heavy breathing) reduces
the BOLD effect dramatically, and that the signal amplitude is dependent on the degree of
cerebrovascular reserve and age. Less well documented are the effects of, for example, mild
hypoglycemia (which reduces the BOLD signal to 60%2), variations in blood pressure or
autoregulation, and the presence of common vasoactive pharmacological agents such as nicotine
and caffeine, or hormonal levels such as estrogen. These factors modulate the BOLD signal and
contribute to the variance in results reported within and between subjects. In addition, the
relationship between blood volume, blood flow and oxygen extraction is not a constant. For
example, even brief periods of hypercapnia have been shown to affect the coupling of the BOLD
signal and flow.3
In infants, the BOLD effect is often reversed compared to adults,4 suggesting that the balance of
physiological variables is different and changes with development. At a more fundamental level,
we do not even know how to interpret BOLD signals in terms of net activation—whether, for
example, inhibitory inputs to a region contribute to the metabolic demand (and hence the BOLD
signal) in the same way as excitatory elements. These issues are tractable research questions that
deserve further attention. Understanding the physiological factors that affect the BOLD signal
will provide a better basis for interpreting the magnitudes of activation, the differences between
conditions, and avoiding possible confounding conditions that may obfuscate the effects of true
differences in behavior.
Rebuttal
An empirical approach to applying new discoveries in medicine may indeed reap benefits. There
are, in fact, many examples of developments that are useful but not totally understood.
Nevertheless, reaching the full potential of new discoveries is often limited by an incomplete
understanding of the underlying science. Cavalier adoption of new methods without such
understanding can be dangerous. For example, the primary impact of aspirin on health care (in
preventing heart attacks, not in analgesia) was overlooked for almost 100 years because we did
not know how the compound worked. The history of medicine is littered with examples in which
so little was understood at the time of use of new drugs and treatments that major harm was
inflicted on innocent people. Thalidomide is one notorious example of this problem. Another
example is the link between aspirin use and Reye's syndrome, which was not made until quite
recently.
The clinical use of MRI Diffusion-Weighted Imaging is limited in practice by our ignorance,
because we do not know how to interpret the serial changes in MRI signals in terms of cellular
177
edema, reversible ischemia, necrosis, infarction, or other biologically-meaningful descriptions. A

new diagnostic technique has to meet qualitatively different standards than an experimental
treatment that looks promising, especially when there are existing alternatives that are less
glamorous but known to work. It seems highly unlikely that studies of weak correlations in
BOLD signals would be adopted by neurologists as persuasive and specific for a major brain
disorder, especially in a way that would affect the patient's treatment or diagnosis, until it has
been established that such variations reflect neural activity, and that their evolution directly
implicates a disease-relevant biochemical or structural process in the brain. That is, these
correlations will probably remain curious observations until we understand more about
fundamental mechanisms.
We know enough already to be cautious about the sensitivity and accuracy of BOLD imaging
and the inherent subjectivity of functional data in some studies. FMRI already provides ample
opportunities for making significant errors in judging the scale and locations of activated cortex.
To reach beyond simple mapping of sensory areas to a more widespread impact in detecting,
diagnosing and treating brain disorders will require "the creation of a reliable tool and
appropriate training of users," as my colleague emphasizes. In this context, however, reliability
will demand understanding the biological and physical factors that affect the results, and training
will require understanding the relationship between indirect BOLD measurements, neural activity
and human cognition.
AGAINST THE PROPOSITION: Robert W. Prost, Ph.D.
Opening Statement
Many phenomena within functional magnetic resonance imaging by the BOLD contrast
mechanism are not yet understood. These phenomena include the steps coupling increased
neuronal activity to the dilation of the end arterioles that control flow in the capillaries.5
Despite an incomplete understanding of the mechanisms involved, the clinical use of fMRI is
expanding rapidly. A growing number of clinical problems are amenable to investigation, and
more importantly, diagnosis by fMRI. One example is the presurgical localization of functions
within the eloquent cortex relative to a lesion. The prediction and prevention of postsurgical
neurological deficit has proven to be of great benefit for both patients and surgeons.
The history of medicine is replete with examples of breakthrough treatments in which the
mechanisms of action are poorly understood, if at all. The most obvious is aspirin. Salicylate, an
extract of willow bark, was described by Hippocrates in 400BC. A synthetic salicylate,
acetylsalylic acid, was first commercially produced by Bayer Pharmaceuticals in 1899. It is still
widely used. The inhibition of cyclo-oxygenase was not discovered until 1990 to be the cause of
the analgesic/anti-inflammatory effect of aspirin. Discovery of the mechanism of action in
acetylaslylic acid did not change the utilization rate of aspirin. The use of aspirin as an
antiplatelet, anticlotting drug was also discovered during the course of treating hemophilic
children with aspirin.6
Diffusion-weighted MRI (DWI) is evolving in a similar manner. Changes in images of the mean
diffusivity of water in the brain have been used to diagnose and determine the spatial extent of
infarct.7 New studies have cast doubt on the mechanisms originally proposed for the observed
178
changes. However, the absence of consensus on mechanisms has not detracted nor derailed the
clinical use of DWI.
Extensions of fMRI to other clinical problems will occur largely independently of an increased
understanding of underlying mechanisms. An example is the concept of resting oscillations in the
BOLD effect, which have been detected by fMRI.8 Li et al. have described the loss of coherence
in these resting state oscillations associated with early stages of Alzheimer's dementia. 9
Mechanisms underlying the loss of coherence are not yet understood. Development of a tool for
early AD diagnosis will likely not depend on the answer.
None of this is to say that fMRI will not suffer misuse. The path to a stable clinical test is
typically littered with the burned-out remains of early adopters, abusers and discarders. Greater
utilization and stability/reliability are dependent on the availability of integrated packages with
stable quality assurance methods and appropriate phantoms. They will also depend on training
radiologists to read and interpret fMRI studies. The rate-limiting step in the clinical utilization of
fMRI is not an understanding of mechanisms, but the creation of a reliable tool and appropriate
training of users.
Rebuttal
While the amplitude of the BOLD signal may not be a linear function of the intensity of the
stimulus, this fact does not preclude the use of fMRI in the clinical setting. Intensity of the
stimulus has been shown to correlate with the number of activated voxels. In bilingual subjects,
the number of activated voxels is always greater in the non-native language.10 Our recent work
has shown the number of activated voxels in the auditory cortex increases with the intensity of
pure tones.11 This would be expected, based on the mechanism of signal transduction in the hair
cells of the inner ear. The firing rate of hair cells increases linearly with increasing tone intensity
between 20 and 40 dB. Above 40 dB, the firing rate of an individual hair cell is saturated.
Increasing tone intensity above 40 dB is encoded by increased recruitment of, rather than
increased response of, individual neurons.
A further barrier to understanding BOLD signal amplitude is the manner in which neuronal
activity is coupled to the elements of metabolic support for the neuron. Studies of glucose and
oxygen utilization show that vasodilatation of the capillary venules always exceeds the increased
metabolic demand of the stimulated neurons.5 The amplitude of the BOLD response may signify
nothing more than the effect of postneuronal activation factors on the over-reactive vascular
response.
An improved understanding of the BOLD response will be an important factor in broadening the
role of fMRI in the clinical domain. However, like so much of medicine, applications of fMRI
will not await a complete understanding of underlying mechanisms. In the interim, fMRI has
already become a valuable clinical tool. The time to start using it is now.
REFERENCES
1. M. Robson, J. L. Dorosz, and J. C. Gore, "Measurements of the temporal fMRI response

of the human auditory cortex to trains of tones," Neuroimage 7, 185–198 (1998).
2. R. P. Kennan, R. J. Jacob, R. S. Sherwin, and J. C. Gore, "Effects of hypoglycemia on
fMRI response to median nerve stimulation in the rat brain," J. Cereb. Blood Flow Metab. 20,
1352–1359 (2000).
179
3. M. V. Dutka, B. E. Scanley, M. D. Does, and J. C. Gore, "Changes in CBF-BOLD

coupling detected by MRI during and after repeated transient hypercapnia in rat," Magn. Reson.
Med. 48, 262–270 (2002).
4. A. W. Anderson, R. Marois, E. R. Colson, B. S. Peterson, R. A. Ehrenkranz, M.
Konstantino, H. Sarofin, K. C. Schneider, J. C. Gore, and L. R. Ment, "Neonatal auditory
activation detected by functional magnetic resonance imaging," Magn. Reson. Imaging 19, 1–5
(2001).
5. C. T. Moonen, Functional Magnetic Resonance Imaging (Springer Verlag, New York,
2000).
6. A. J. Quick, "Aspirin, platelets and hemostasis," N. Engl. J. Med. 283, 598 (1970).
7. A. Gass, T. Niendorf, and J. G. Hirsch, "Acute and chronic changes of the apparent
diffusion coefficient in neurological disorders—biophysical mechanisms and possible underlying
histopathology," J. Neurol. Sci. 186, S15–23 (2001).
8. V. Haughton and B. Biswal, "Clinical application of basal regional cerebral blood flow
fluctuation measurements by FMRI," Adv. Exp. Med. Biol. 454, 583–590 (1998).
9. S. J. Li, Z. Li, G. Wu, M. J. Zhang, M. Franczak, and P. G. Antuono, "Alzheimer disease:
Evaluation of a functional MR imaging index as a marker," Radiology 225, 253–259 (2002).
10. T. Nakai et al., "A functional magnetic resonance imaging study of listening
comprehension of languages in human at 3 Tesla-comprehension level and activation of the
language areas," Neurosci. Lett. 263(1), 33–36 (1999).
11. K. J. Lasota, J. L. Ulmer, J. B. Firszt, B. B. Biswal, D. L. Daniels, and R. W. Prost,
"Intensity-dependent activation of the primary auditory cortex in functional magnetic resonance
imaging," J. Comput. Assist. Tomogr. 27(2), 213–218 (2003).
180
4.3. Simultaneous PET/MR will replace PET/CT as the molecular

multimodality imaging platform of choice
Habib Zaidi and Osama Mawlawi
OVERVIEW
With the combination of PET and CT images in dual-modality PET/CT units, it is now possible
to accurately align the functional information obtained with PET to the anatomical structures
revealed by CT. This is a significant improvement over previous methods of combining these
two modalities by ―fusing‖ images obtained in sequential studies, with all the problems
associated with precise patient positioning. The oncological community has so embraced this
new technology that PET/CT units are now becoming commonplace. Indeed, PET units are now
rarely purchased without being combined with CT. However, many would argue that the
anatomical data derived from CT is not as complete as that which could be obtained with MRI,
and the metabolic information that can be obtained with PET is somewhat limited compared with
that which might be obtained with magnetic resonance, especially with functional MRI (fMRI)
and MR spectroscopy (MRS). This has led to the recent development of combined PET/MR
units, which are being promoted as even better than PET/CT. The premise that PET/MR will
replace PET/CT as the molecular multimodality imaging platform of choice is the topic debated
in this month's Point/Counterpoint.
Arguing for the Proposition is Habib Zaidi, Ph.D. Dr. Zaidi received a Ph.D and Habilitation
(Privat-docent), both in Medical Physics, from the University of Geneva. He is senior physicist
and head of the PET Instrumentation & Neuroimaging Laboratory at Geneva University Hospital,
where he is actively involved in the development of imaging solutions for cutting-edge
interdisciplinary biomedical research and clinical diagnosis. He is a member of the editorial
board and/or serves as scientific reviewer for several scientific journals. He is a senior member of
the IEEE and Vice Chair of the Professional Relations Committee of the IOMP. He is involved in
the evaluation of research proposals for European and international granting organizations and
participates in the organization of international symposia and conferences. He is a recipient of
many awards and distinctions and has been an invited speaker of many keynote lectures at an
international level.
Arguing against the Proposition is Osama Mawlawi, Ph.D. Dr. Mawlawi received his Ph.D in
Biomedical Engineering from Columbia University in NY. He did his graduate training in PET
imaging at Memorial Sloan Kettering Cancer Center in New York City before accepting a joint
faculty position in the Departments of Radiology and Psychiatry at Columbia University Medical
Center, where he focused on neuroreceptor imaging using PET. He is currently an Associate
Professor of Imaging Physics at M.D. Anderson Cancer Center in Houston, Texas, and is the lead
PET/CT physicist at the center. Dr. Mawlawi is a reviewer for numerous international journals,
has been an invited speaker at many national and international conferences, and is the recipient
of several grants from public and private sources.
FOR THE PROPOSITION: Habib Zaidi, Ph.D.

181
Opening Statement
The recent introduction of PET/MR technology is considered by many experts to be a major

breakthrough that will potentially lead to a paradigm shift in healthcare and revolutionize clinical
practice. Several research groups in academic and corporate settings are focusing on the
development of various configurations of MR-compatible PET inserts to allow simultaneous
scanning using the most highly sophisticated molecular imaging technologies available
today.1,2,3,4,5,6 Compared to PET/CT, where sequential scanning was (erroneously) considered to
be the ultimate solution for image coregistration to correlate structural and functional information
thus allowing anatomic localization of abnormal tracer uptake or facilitating the process of
differentiating normal from abnormal uptake, simultaneous PET/MR has many additional
features. First, for small animal studies, simultaneous scanning reduces time under anesthesia and
enables scanning under identical physiological conditions. Second, high-field MRI generates
high resolution anatomical and structural images offering better soft-tissue contrast resolution
and a large variety of tissue contrasts compared to CT, and allows for functional MRI, thus
enabling temporal correlation of blood flow with metabolism or receptor expression in brain
studies and, more importantly, is capable of assessing flow, diffusion, perfusion, and cardiac
motion in one single examination. Third, MRI can be combined with MRS to measure spatially
matched regional biochemical content and to assess metabolic status or the presence of neoplasia
and other diseases in specific tissue regions. Finally, MRI does not use any ionizing radiation and
therefore can be used without restrictions in serial studies, for pediatric cases, and in many other
situations where radiation exposure is a concern.
A major advantage cited for PET/CT is that it enables a reduction in the overall scanning time by
using CT images for attenuation correction. However, it does this at the expense of a substantial
increase in absorbed dose, a significant issue when scanning normal subjects and small animals,
as it might change the animal model being studied.7 In comparison to CT, MRI typically is more
expensive, involves longer scan times, and produces anatomical images from which it is more
difficult to derive maps for attenuation correction of the PET emission data. However, some
solutions do exist as demonstrated by a proof of concept for using segmented MRI-guided
attenuation compensation in brain PET,8 and plenty of opportunities remain for creative advances
in MRI-guided attenuation correction in whole-body PET imaging.
Whereas many technical problems have been recently solved, it is recognized that
implementation and operation of a combined PET/MR system is still facing many important
challenges that must be overcome through research. Many design configurations based on the use
of detector readout technologies insensitive to magnetic fields have been proposed, including
avalanche photodiodes and, more recently, silicon photomultipliers, particularly for preclinical
systems. Moreover, one of the major vendors recently married a PET insert to a 3T MR head
scanner,4 making its application to humans in research settings (and possibly extension to clinical
whole-body PET/MR) a near certainty as prototypes are being deployed in some European
institutions.
Weighing the advantages and drawbacks of each technology renders the conclusion
―simultaneous PET/MR will replace PET/CT as the molecular multimodality imaging platform of
choice‖ not only plausible but also obvious. This technology will likely succeed in unifying the
four promising molecular imaging techniques PET, structural MRI, fMRI, and MRS, which is in
sharp contrast to the limited information provided by dual-modality PET/CT imaging.
182
AGAINST THE PROPOSITION: Osama Mawlawi, Ph.D.
Opening Statement
It is always difficult to predict the extent of success a new technology will achieve particularly
when it is still in its infancy as is the case with PET/MR. In general however, for a nascent
technology to successfully replace an established standard, its capabilities should not only
emulate the standard but also constitute a demonstrable advantage that undeniably leads to its
widespread use, while its disadvantages should not offset its potential benefits. In this regard I
will argue, for now, that PET/MR will not replace PET/CT as the modality of choice for
molecular imaging at least from a clinical evaluation standpoint. For research applications on the
other hand, much can be said in favor of PET/MR.
PET/MR has the capability of emulating the achievements already established by PET/CT. Scan
duration with PET/MR is anticipated to be similar to PET/CT or slightly longer depending on the
pulse sequence used.9 This can, however, only be achieved if the design of PET/MR allows for
concurrent rather than sequential data acquisition, as is the case with PET/CT. Furthermore, the
instantaneous fusion of anatomical and functional data, which facilitated the acceptance of
PET/CT, can be accomplished with PET/MR9,10,11 and the use of MR for attenuation correction,
although challenging, is also presumably feasible as has been shown at least for brain imaging. 8
However, as mentioned earlier, for PET/MR to replace PET/CT it should, in addition to
emulating the advantages of PET/CT, provide a clinically practical and tangible advantage in
image acquisition, interpretation, and diagnosis.
PET/CT is currently mainly used for whole-body oncological evaluations, an application

comprising the majority of reimbursable indications for PET/CT. 12 In this regard, for PET/MR to
replace PET/CT it should be able to provide a better alternative particularly in whole body
imaging. However, since the pairing of MR with PET is not envisioned to further improve the
PET imaging portion over what has already been accomplished with CT but rather to augment it
with functional (fMRI) or spectroscopic (MRS) data, both of which are not routinely used for
diagnostic evaluation, the question then becomes: Is it better from a diagnostic perspective to pair
a PET scanner with CT or MR? The answer to this question largely depends on the application,
ease of use, and cost. Since its introduction in the early 1980s, MR has steadily gained
acceptance, but it has not replaced CT in many areas, mainly because of the reasons above. The
same applies to PET/MR replacing PET/CT in routine clinical imaging. Until the diagnostic
advantages of whole body PET/MR are well established, reimbursable, and supersede any
additional diagnostic advantages that PET/CT might introduce, I believe that PET/MR will be
restricted to research and will have difficulty replacing the widely accepted use of PET/CT.
One area where PET/MR has a clear advantage over PET/CT is lower patient radiation exposure.
This advantage should by itself be sufficient to predict that PET/MR will replace PET/CT.
Unfortunately, however, we have repeatedly seen that efforts for dose reduction are
circumvented by other dominating factors such as cost, speed, and ease of use13,14 suggesting that
this reasoning might not be strong enough to induce the suggested change in PET imaging.
In summary, PET/MR scanners have a lot of potential advantages. However, for these
advantages to assist PET/MR in replacing PET/CT in routine clinical evaluation, they first have
to be cost effective and easy to realize and, most importantly, become a necessary component of
the routine whole body diagnostic evaluation of cancer patients using PET imaging.
183
Rebuttal: Habib Zaidi, Ph.D.
I concur with my colleague that current PET/CT technology is more comfortable for end-users
and that the use of PET/MR in clinical and research settings requires extensive technical and
organizational efforts that may restrict its use in the short term to academic centers having the
required scientific resources. I also agree that any new technology should be assessed carefully
with respect to benefits conveyed to patients before widespread acceptance and adoption. We
clearly need large-scale studies to demonstrate the clinical benefits of PET/MR and, more
importantly, to define where PET/CT is sufficient and where PET/MR is needed. However, I
disagree with the main arguments raised to claim that PET/CT is, and will remain, the clinical
standard in the foreseeable future. First, combined PET/MR is also aiming at improving the
diagnostic relevance of the PET imaging portion but in a more profound way, which is not
limited to providing anatomical information for mapping of metabolic abnormalities and
shortening transmission scanning time, as PET/CT does. In addition to offering a diversity of
tissue contrasts, MR will provide a wealth of additional information through fMRI and MRS to
enhance the diagnostic performance and quantitative capabilities of PET. More importantly,
using simultaneous (rather than sequential) scanning will resolve many of the impediments to
precise coregistration of anatomo-molecular information and accurate attenuation correction.
Second, reimbursement issues are mainly driven by prospective clinical studies that demonstrate
improvements in health outcomes conveyed by an imaging modality for a given indication.
Therefore, given the higher soft tissue contrast resolution of MRI and its highest sensitivity and
specificity for many indications (e.g., detection of liver metastases), 9 coverage for PET scans will
undoubtedly be expanded.
This having been said, I would like to challenge my colleague further by asking: ―Is PET/CT
unanimously recognized as the standard imaging technology for clinical oncology?‖ One should
bear in mind that this is still a controversial issue, since many investigators claim that it has a
limited role in many indications including lymphomas, lung nodules, and brain tumors.15 Time
will dictate whether PET/MR will influence the standard for future PET instrumentation, which
is poised to advance molecular imaging and influence clinical and research practice.
Rebuttal: Osama Mawlawi, Ph.D.
I agree with Dr. Zaidi that PET/MR has many potential advantages, provided that current
technological barriers facing its development are resolved. As I mentioned in my opening
statement, PET/MR theoretically provides the same advantages that PET/CT provided to
dedicated whole body PET imaging such as anatomical landmarks, shorter scan durations, and
attenuation correction. In addition, PET/MR potentially can augment PET imaging with a wealth
of other information such as functional and spectroscopic data that may be helpful in improving
patient management as well as understanding tumor biology. Furthermore, with PET/MR this
information is obtained at a reduced patient radiation exposure compared to PET/CT. So in
essence, pairing PET with MR can only provide an added advantage over PET/CT. However,
going as far as suggesting that PET/MR will replace PET/CT might be premature at this stage,
particularly since none of the suggested added advantages of PET/MR have been shown to be
clinically justified, practical and, most importantly, cost effective in routine whole body
oncological imaging.
There is no doubt that in order to assess the need for PET/MR in a clinical setting, such a hybrid
modality should be made available at least in large research centers. Results from studies
conducted on these systems will then provide the necessary data to justify their routine clinical
184
use and eventually convince the medical community about the merits and cost effectiveness of
PET/MR. Until that happens, I believe that PET/CT will continue to be the modality of choice in
whole body oncological imaging.
REFERENCES
1. Y. Shao, S. R. Cherry, K. Farahani, and K. Meadors, ―Simultaneous PET and MR

imaging,‖ Phys. Med. Biol. 42, 1965–1970 (1997).
2. J. E. Mackewn et al., ―Design and development of an MR-compatible PET scanner for
imaging small animals,‖ IEEE Trans. Nucl. Sci. 52, 1376–1380 (2005).
3. B. J. Pichler et al., ―Performance test of an LSO-APD detector in a 7-T MRI scanner for
simultaneous PET/MRI,‖ J. Nucl. Med. 47, 639–647 (2006).
4. Z. Burbar et al., ―PET performance of MR/PET brain insert tomograph,‖ Proc. IEEE
Nuclear Science Symposium and Medical Imaging Conference, San Diego, CA (2006).
5. A. J. Lucas et al., ―Development of a combined microPET-MR system,‖ Technol. Cancer
Res. Treat. 5, 337–341 (2006).
6. M. S. Judenhofer et al., ―Simultaneous PET/MR images, acquired with a compact MRI
compatible PET detector in a 7 Tesla magnet,‖ Radiology (in press).
7. R. Taschereau, P. L. Chow, and A. F. Chatziioannou, ―Monte Carlo simulations of dose
from microCT imaging procedures in a realistic mouse phantom,‖ Med. Phys. 33, 216–224
(2006).
8. H. Zaidi, M.-L. Montandon, and D. O. Slosman, ―Magnetic resonance imaging-guided
attenuation and scatter corrections in three-dimensional brain positron emission tomography,‖
Med. Phys. 30, 937–948 (2003).
9. M. D. Seemann, ―Whole-body PET/MRI: the future in oncological imaging,‖ Technol.
Cancer Res. Treat. 4, 577–582 (2005).
10. P. K. Marsden, D. Strul, S. F. Keevil, S. C. Williams, and D. Cash, ―Simultaneous PET
and NMR,‖ Br. J. Radiol. 75, S53–59 (2002).
11. C. Catana et al., ―Simultaneous acquisition of multislice PET and MR images: Initial
results with a MR-compatible PET scanner,‖ J. Nucl. Med. 47, 1968–1976 (2006).
12. M. Beebe, J. Dalton, M. Espronceda, D. Evans, and R. Glenn, CPT 2007 Professional
Edition (American Medical Association Press, 2006).
13. F. A. Mettler, Jr., P. W. Wiest, J. A. Locken, and C. A. Kelsey, ―CT scanning: patterns of
use and dose,‖ J. Radiol. Prot. 20, 353–359 (2000).
14. J. M. Boone, ―Multidetector CT: opportunities, challenges, and concerns associated with
scanners with 64 or more detector rows,‖ Radiology 241, 334–337 (2006).
15. A. Alavi, A. Mavi, S. Basu, and A. Fischman, ―Is PET-CT the only option?,‖ Eur. J.
Nucl. Med. Mol. Imaging (in press).
185
4.4. Methods for image segmentation should be standardized and

calibrated
Edward Chaney and Geoffrey Ibbott
OVERVIEW
Methods of image segmentation and fusion are being widely deployed to characterize three-
dimensional (3D) treatment volumes in intensity modulated radiation therapy (IMRT) and other
sophisticated therapeutic approaches. Some physicists believe that these methods must be
standardized, and that calibration methods must be developed, in order to ensure that the 3D
treatment volumes depicted by the methods are true representations of the volumes of interest in
the patient. Other physicists believe that image segmentation and fusion are evolving
technologies that should be allowed to progress without the constraints of standardization and
calibration. This controversy is the subject of this month's Point/Counterpoint. (Note: the authors
emphasize that the positions on the proposition are taken in the spirit of debate, and do not
necessarily reflect their personal or their employers' viewpoints.)
Arguing for the Proposition is Edward Chaney, Ph.D. Dr. Chaney is a Professor in the
Department of Radiation Oncology at the University of North Carolina (UNC) School of
Medicine, where he was Director of the Physics Division from 1979 until 2003. In 2003 he
retired as Director to devote more time to research and teaching. He is a senior investigator in the
Medical Image Display and Analysis Group at UNC and focuses on computer-based methods for
image analysis with image-guided radiation treatment planning and delivery as primary driving
problems. He is a cofounder of Morphormics, Inc., a startup company specializing in computer
systems for image analysis.
Arguing against the Proposition is Geoffrey Ibbott, Ph.D. Dr. Ibbott is Professor and Chief of the
Section of Outreach Physics at the UT M.D. Anderson Cancer Center in Houston. The section
includes several programs known to many medical physicists, including the Accredited
Dosimetry Calibration Laboratory, Radiation Dosimetry Services, and the Radiological Physics
Center (RPC). As Director of the RPC, Dr. Ibbott has a particular interest in the quality assurance
of cooperative group clinical trials. When not busy with his professional activities, Dr. Ibbott can
be found ballroom dancing with his wife, Diane, or sailing in Galveston Bay.
FOR THE PROPOSITION: Edward Chaney, Ph.D.
Opening Statement
The accuracy and reproducibility of radiation therapy can be improved by standardizing and
calibrating image segmentation methods. Segmentation is a commonly performed procedure that
affects critical treatment planning and delivery decisions. Image-guided 3D and emerging four-
dimensional (4D) planning and delivery methods require one or more user-created models of the
patient to localize and display objects of interest, position beams, and shape beam apertures,
186
compute DVHs and volume-weighted metrics, characterize temporal changes in patient anatomy,
and transfer information from one or more reference images to inter- and intra-treatment images
for accurate and reproducible targeting. The structures comprising the patient model are defined
by segmenting volume images. Due to the large number of departments practicing image-guided
planning and delivery, it is likely that segmentation is performed more often as a clinical
procedure in radiation oncology than for all the other medical specialties combined.
Calibrating a dose to a point in water has served well the two–dimensional (2D) and
nonconformal 3D eras but is insufficient for the modern era, which is distinguished by tight
margins and steep dose gradients intended to shrink-wrap the high-dose region around the target
while conformally avoiding nearby normal tissues. The modern approach, particularly in
conjunction with inverse planning methods, is exquisitely sensitive to geometric variations in the
patient model. The quality of manual segmentation is degraded by user-specific systematic and
random intra- and inter-user variabilities that in turn are manifested as suboptimal plans and
imprecise targeting. Emerging automatic methods1,2,3,4,5,6,7,8 promise to significantly reduce
random variabilities, leaving predominantly systematic errors that in principle are correctable by
training, algorithm "tuning," or supervised editing. While it would be impractical if not
impossible to calibrate human segmentation, it is possible to standardize and calibrate automatic
methods. Standardization would eliminate nonuniform in-house practices that confound
comparison of clinical studies from different sites and impede accurate export and import of
protocols. Calibration would assure compliance with an accepted standard.
Current automatic methods produce approximate segmentations intended for supervised editing.
As automatic methods improve and gain the confidence of users, and as the number of imaging
procedures increases, it is likely that practice will shift toward minimal supervision and
eventually to almost complete reliance on automation. Minimal supervision elevates the desired
standard of performance and adds urgency to finding standardization and calibration solutions.
Now is the time to begin working on the issues related to finding those solutions. One challenge
to be faced is characterizing the performance of algorithms in clinically relevant terms.
Performance characterization is of interest in computer vision and important questions are being
addressed.9,10,11 Practical issues include agreeing on standard practices and developing calibration
methods that can be widely implemented. The AAPM can take a leadership role by including
focused sessions during annual meetings and forming task groups to study the major issues and
make recommendations. Inaction assures that the promises of standardization and calibration will
be lost.
Standardization refers to developing protocols to encourage uniform segmentation practices. A

protocol for the rectum might specify, for example, the length of rectum to be segmented, and
whether the rectal wall and/or the anterior portion near the prostate should be segmented.
Calibration refers to quantitating segmentation accuracy and reproducibility.
Rebuttal
Dr. Ibbott presents a clear picture of the professional and scientific challenges that must be
addressed to find widely acceptable solutions for calibration and standardization of image
segmentation for radiation therapy. Standardization of target volumes is particularly problematic
and his point is well made that better understanding and consensus are needed. This argument
applies to organs at risk (OAR) as well, but the imaging issues are not as complex. Also in the
scientific arena, further research is needed to develop, validate, and calibrate segmentation
methods. Traveling the road to understanding and consensus is a community venture that
187
requires compiling and vetting information, education, airing ideas, and open debate.
Professional and scientific organizations such as the AAPM can play an important role in this
process.
AGAINST THE PROPOSITION: Geoffrey Ibbott, Ph.D.
Opening Statement
In nearly every aspect, standardization in radiation oncology is a desirable objective. It is widely

agreed, for example, that treatment machine calibration must be traceable to national standards,
to assure that dose delivery is consistent among departments. 12 Data communication, particularly
that involving radiological images and radiation therapy treatment plans, must be conducted
according to protocols such as DICOM to ensure that the data are received intact. Seemingly
obvious things such as the units and coordinate systems used for describing radiation beam size
and orientation must be standardized to facilitate the accurate communication of these
parameters.13 Several unfortunate accidents in radiation therapy have occurred because of
misunderstandings of the coordinate system used to describe the placement of the radiation field,
or confusion over the units of a calibration coefficient. 14
With regard to the definition of target volumes and OAR in radiation therapy, it would seem that
standardization would have many benefits. Defining target volumes according to a standard
could streamline operations in a radiation therapy department, and might result in time and cost
savings. Perhaps even more significantly, standardizing target volume definition processes could
improve the quality of clinical trials by ensuring that patients receive equivalent care at multiple
institutions. One day, we will probably reach this point.
Today, however, may be too soon to develop standards for defining target volumes and OAR.
Numerous studies have shown that physicians rarely agree on the shape and size of a target
volume.15 Physicians involved in clinical trials suggest that inconsistent identification of target
volumes is probably a greater cause of variations in patient treatment than is the implementation
of new treatment technologies.16 There is no "gold standard" for any target volume or other
structure. In fact, in a recently closed multi-institutional study of IMRT for treatment of the
oropharynx, the principal investigator revised the contours drawn by physicians registering
patients in the trial, to assure that dose-volume histograms were calculated according to his
criteria.17 Another RTOG/NSABP trial that opened recently requires that participants
demonstrate their willingness and ability to define the target volume according to the principal
investigators' criteria.18 It is unlikely that these PIs would have accepted a standards agency's
definition of the targets and other structures.
The expansion into radiation therapy planning of imaging technologies such as MR and PET is
changing the way target volumes are defined. On the horizon are molecular imaging techniques
that promise to allow routine identification of tumor-bearing tissue in organs such as the prostate,
further improving the radiation oncologist's ability to define the target. In addition, the current
interest in 4D imaging and treatment to accommodate respiratory motion raises new uncertainties
in the identification of target volumes. Standardizing now could limit the development of these
techniques and discourage new research.
188
Until the identification of tumor volumes is better understood and consensus is achieved in the
radiation therapy community, the standardization of target volume segmentation should remain a
research project.
Rebuttal
In his opening statement, my opponent claims that standardization and calibration of

segmentation methods can improve the accuracy and reproducibility of radiation therapy. It is
hard for me to disagree with this claim. As one whose career has been intimately associated with
the development of standards,13 and the supervision of calibration laboratories,12 I support both
activities. But I question whether automatic segmentation can improve the quality of radiation
therapy. Is it true, as it is with dosimetry, that complying with a consistent standard is more
important than that the standard be correct?
With regard to defining target volumes, being consistently in error is not acceptable. Removing
random variabilities could be detrimental to patient care, if the wrong standard is chosen.
Dr. Chaney says that calibration of segmentation techniques would assure compliance with an
accepted standard. If only there were such a standard! The literature is filled with examples of the
disagreements encountered when one asks several physicians to define target volumes15,19,20
There are also examples of the disagreements demonstrated when different imaging modalities,
such as CT and MRI, are used for the segmentation procedure.21 The introduction of PET has
improved the ability to identify the location and extent of lung tumors, leading radiologists to
change the volumes delineated on CT images. 22
My opponent points to a recent publication discussing the development and design of a

segmentation algorithm.3 In a related paper, the algorithm is used to perform segmentation of
kidneys from CT images.23 While the performance of the algorithm is impressive, the authors
note that significant disagreements occurred between two experienced physicians who contoured
the same kidney. The disagreement between the automatic segmentations and those drawn by
humans was similar. It is not clear how well this algorithm might function with structures that are
less clearly defined than the kidney, which is the case for many tumors.
Dr. Chaney concludes that the challenges facing the implementation of segmentation algorithms
include finding ways to characterize their performance, developing calibration methods, and
agreeing on standard practices. Developing the consensus needed to address these issues could be
supported by the AAPM. I wholeheartedly agree, and look forward to progress in this area. In the
meantime, as was stated before, I believe automatic segmentation should remain in the research
arena.
REFERENCES
1. Model-Based Analysis of Medical Images, in special issue of IEEE Trans. Med. Imaging
18(10) (1999).
2. T. McInerney and D. Terzopoulos, "Deformable models in medical image analysis: a
survey," Med. Image Anal 1, 91–108 (1996).
3. S. M. Pizer, P. T. Fletcher, S. Joshi, G. Gash, J. Stough, A. Thall, G. Tracton, and E. L.
Chaney, "A method and software for segmentation of anatomic object ensembles by deformable
M-Reps," Med. Phys. 32, 1335–1345 (2005).
189
4. G.S. Mageras, S. Joshi, B. Davis, A. Pevsner, A. Hertanto, E. Yorke, K. Rosenzweig, and

C. Ling, "Evaluation of an automated deformable matching method for quantifying lung tumor
motion in respiration-correlated CT images," in Proceedings of the XIVth International
Conference on the Use of Computers in Radiation Therapy (ICCR, Seoul, 2004),
http://www.iccr.info/paper_4.asp.
5. D. Freedman, R. Radke, T. Zhang, Y. Jeong, M. Lovelock, and G. Chen, "Model-based
segmentation of medical imagery by matching distributions," IEEE Trans. Med. Imaging 24,
281–292 (2005).
6. W. Lu, M.-L. Chen, G. Olivera, K. Ruchala, and T. Mackie, "Fast free-form deformable
registration via calculus of variations," Phys. Med. Biol. 49, 3067–3087 (2004).
7. V. Pekar, T. R. McNutt, and M. R. Kaus, "Automated model-based organ delineation for
radiotherapy planning in prostatic region," Int. J. Radiat. Oncol., Biol., Phys. 60, 973–980
(2004).
8. H. Wang, L. Dong, M. Lii, A. Lee, R. De Crevoisier, R. Mohan, J. Cox, D. Kuban, and
R. Cheung, "Implementation and validation of a three-dimensional deformable registration
algorithm for targeted prostate cancer radiotherapy," Int. J. Radiat. Oncol., Biol., Phys. 61, 725–
735 (2005).
9. Performance Characterization in Computer Vision, Computational Imaging and Vision
Series Vol. 17, edited by R. Klette, H. S. Stiehl, M. A. Viergever, and K. L. Vincken (Springer,
Berlin, 2000).
10. Project on Performance Characterization in Computer Vision, Pilot European Image
Processing Archive; http://peipa.essex.ac.uk/benchmark/
11. S. Warfield, K. Zou, and W. Wells, "Simultaneous Truth and Performance Level
Estimation (STAPLE): An algorithm for the validation of image segmentation," IEEE Trans.
Med. Imaging 23, 903–920 (2004).
12. G. S. Ibbott, F. H. Attix, T. W. Slowey, D. P. Fontenla, and M. L. Rozenfeld,
"Uncertainty of calibrations at the accredited Dosimetry Calibration Laboratories," Med. Phys.
24, 1249–1254 (1997).
13. IEC 61217: Radiotherapy Equipment—Coordinates, Movements and Scales. Ed. 1,
Amendment 1 (International Electrotechnical Commission, Geneva, 2002).
14. Lessons learned from accidental exposures in radiotherapy, Technical Reports Series No.
17 (International Atomic Energy Agency, Vienna, 2000).
15. P. Tai et al., "Variability of target volume delineation in cervical esophageal cancer," Int.
16. E. Glatstein, "The return of the snake-oil salesmen," Int. J. Radiat. Oncol., Biol., Phys.
55, 561–562 (2003).
17. RTOG 0022: "Phase I/II study of conformal and intensity modulated irradiation for
oropharyngeal cancer" (Radiation Therapy Oncology Group, Philadelphia, 2001).
18. NSABP B-39/RTOG 0413, "A randomized phase III study of conventional whole breast
irradiation (WBI) versus partial breast irradiation (PBI) for women with stage 0, I, or II breast
cancer" (National Surgical Adjuvant Breast and Bowel Project, Pittsburgh, and Radiation
Therapy Oncology Group, Philadelphia, 2005).
19. R. J. Dowsett et al., "Contouring structures for 3-dimensional treatment planning," Int. J.
20. G. Leunens et al., "Quality assessment of medical decision making in radiation oncology:
Variability in target volume delineation for brain tumors," Radiother. Oncol. 29, 169–175
(1993).
21. C. Rasch et al., "Definition of the prostate in CT and MRI: A multi-observer study," Int.
190
22. J. Bradley et al., "Impact of FDG-PET on radiation therapy volume delineation in non-
small-cell lung cancer," Int. J. Radiat. Oncol., Biol., Phys. 59, 78–86 (2004).
23. M. Rao et al., "Comparison of human and automatic segmentations of kidneys from CT
images," Int. J. Radiat. Oncol., Biol., Phys. 61, 954–960 (2005).
191
4.5. Standards for image quality and radiation exposure impede the
pursuit of optimized quality/dose ratios in radiology
David Goodenough and Caridad Borrás
Reproduced from Medical Physics, Vol. 27, No. 2, pp. 273-275, February 2000
OVERVIEW
Universal standards for radiation exposure and image quality have long been objectives of
numerous ad hoc task groups, various government agencies and national and international
organizations, and committees of the AAPM and many other scientific and professional
organizations. Efforts directed to attaining these objectives or measuring compliance with them
have occupied the time and effort of almost every medical physicist involved in medical
imaging. However, these efforts may be antithetic to the optimization of quality/ dose ratios in
medical imaging, because they establish intermediate goals for physicists that divert their search
for the best combinations of image quality and patient dose. One of the first rules of Total
Quality Improvement (TQI), according to its developer W. Edwards Deming, is to do away with
all standards so as not to impede the commitment to continuous improvement of quality. The
apparent conflict between development of exposure and quality standards, and principles such as
TQI and Total Quality Management, is explored in this issue of Point/Counterpoint.
Arguing for the Proposition is David Goodenough, Ph.D. Dr. Goodenough is Professor of
Radiology and Co-Director of The Institute for Medical Imaging and Image Analysis at The
George Washington University. He received his B.S. in Physics in 1967, and his Ph.D. in
Medical Physics in 1972 from the University of Chicago where he served as a member of the
Radiology Faculty and Center for Radiologic Image Research. In 1974 he served as Visiting
Associate with the Bureau of Radiological Health (FDA) before serving as Assistant Professor of
Radiology at Johns Hopkins University. In 1975 he moved to GWU. Dr. Goodenough is
interested in CT, MRI, and SPECT as well as PACS and teleradiology systems, ROC curves and
efficacy studies.
Arguing against the Proposition is Caridad Borrás, D.Sc. Cari Borrás came from Spain to the
United States in 1966 as a Fulbright-Hays scholar. In 1974 she received a Doctoral of Sciences
Degree from the University of Barcelona, Spain, following completion of a thesis carried out at
Thomas Jefferson University, Philadelphia. She then worked at the West Coast Cancer
Foundation in San Francisco. Since 1988 she has been the Regional Adviser in Radiological
Health at the Pan American/World Health Organization in Washington, DC, where she advises
the Ministries of Health of the Americas concerning the development and adoption of
Radiological Standards. She is an ABR examiner and a Fellow of the ACR.
FOR THE PROPOSITION: David Goodenough, Ph.D.
Opening Statement
Standards for Image Quality and Radiation Exposure may sometimes be necessary, but they
should be a course of last resort. The need for standards may often mean that the prevailing
educational and common sense approaches have failed. In particular, such standards may set a
192
‗‗lower bar‘‘ of performance rather than lead to an ever-increasing search for improvement and
optimization.
Standards also tend to breed bureaucracies of enforcement sometimes filled by power seeking
individuals with little or no knowledge of the scientific facts. I am sure most of us could identify
scores of agencies and groups considering such issues. Such bureaucracies add cost and
consternation to many well-meaning endeavors.
Standards are particularly doubtful in a complex multivariable paradigm, such as the relationship
between Dose and Image Quality. One variable may legitimately trade-off with another, e.g.,
increased resolution may lead to increased noise; or increased dose may be necessary for lower
noise? Then too, Dose regulating bodies seem fond of imposing an upper limit on dose. Why not
a lower limit? It could be argued that the worst radiation burden is Dose that cannot supply the
required diagnostic information, as for example, an underexposed film. Moreover, consider the
legitimate choices of image receptor sensitivity, resolution, film speed, grid ratio (or no grid).
Speed differences range over orders of magnitude. Can anyone who is not thoroughly familiar
with the scientific data monitor and regulate this choice? In fact, regulators would need thorough
medical knowledge of the kinds of signals to be detected. It is probable that most individuals
would far prefer to bear the risk of added radiation burden if it meant the difference between
finding or missing a small life-threatening lesion.
Another important issue in regulation of image quality and radiation exposure is the
differentiation between screening versus indicated studies. Clearly, a known cancer induction
risk needs to be balanced against diagnostic efficacy of a screening device (such as
mammography). In that sense, dose guidelines may be appropriate; however, even here a ‗‗one
Dose limit fits all‘‘ approach does not always make sense. Exceptions may need to be made
based on individual patient factors if it means the difference between detection or miss.
One problem is that standard groups seem to be moving into the arena of indicated procedures,
e.g., CT scanners. It is my belief that the great majority of CT procedures can and should be
performed for indicated reasons. In this case, it seems the benefit/risk ratio is already greatly in
favor of the CT procedure. CT should not generally be used for screening purposes. Over the
years, I have seen various CT vendors promote or criticize competitors‘ Dose levels often out of
envy of their image quality. Likewise, Dose optimization is often a cover for limited power
capacity of x-ray tubes, or limited dynamic range of detectors. There may be many
agendas in play at the same time.
More than standards, we need increased physician education of actual dose levels and the
legitimate dose tradeoffs for Image Quality.
Rebuttal
I admire my counterpart for carefully defining her use of the term ‗‗standard.‘‘ In the sense of
her definition of the term, I am not opposed to the concept of standards when defined as ‗‗precise
criteria‘‘ or as ‗‗definitions.‘‘ Such attributes are the bulwarks of much of scientific
methodology. I am certainly not opposed to ‗‗guidelines.‘‘ I do, however, continue to oppose that
part of the definition that includes ‗‗rules‘‘ that, I believe, need a ‗‗regulatory body‘‘ to enforce.
193
My counterpart notes convincingly, with her example of the 200 vs 400 film speed issue, some
of the many complications that may arise in standardization efforts. These are the kinds of issues
that concern me.
I also agree with my counterpart, that standards for radiation dose and image quality should not
divert the Medical Physicist‘s search for the best combination of both. However, the reality is
that in today‘s managed care environment, the Medical Physicist may be an endangered species.
Administrators may not be receptive to efforts to improve a situation that may involve additional
costs, particularly when the current system ‗‗meets the regulatory guidelines (rules).‘‘ Other than
a polarity change, the ‗‗concern about cost‘‘ gene may be similar in regulators and
administrators.
I do not believe we should follow the European Community‘s compulsion to regulate and
standardize everything, including aspects of dose and image quality. In my opinion, American
Radiology and Medical Physics need broad opportunities for creative excellence, not rigidly
formulated rules or standards that may actually impede exploration of improved techniques.
AGAINST THE PROPOSITION: Caridad Borrás, D.Sc.
Opening Statement
I strongly disagree with the application of W. Edwards Deming‘s concept of standards—mostly

developed as production standards—to image quality/dose ratios.
According to the New Webster dictionary, a standard is ‗‗anything taken by general consent as a
basis of comparison, or established as a criterion.‘‘ According to the International Organization
for Standardization (ISO),1 technical ‗‗standards are documented agreements containing
technical specifications or other precise criteria to be used consistently as rules, guidelines, or
definitions of characteristics to ensure that materials, products, processes and services are fit for
their purpose.‘‘ Based on these concepts, standards for image quality and dose should be
understood as criteria that document the state of the art in image quality and dose at the time of
their publication/adoption. They need to be revised as radiological technology and radiologists‘
training evolve. (ISO revises its standards every five years.)
The problem is that while there are published reference or guidance levels for diagnostic
exposures,2–4 there are few ‗‗quantitative‘‘ standards on image quality, and when they exist, they
are not correlated with dose. For example, FDA‘s Quality Mammography Standards require
phantom images with a ‗‗minimum score‘‘ (a quantity, albeit determined subjectively) and doses
per craniocaudal exam not to exceed 3 mGy, an upper boundary. Correlation between dose and
image quality was attempted in European guidelines on quality criteria, which were defined in
terms of anatomical details seen in a radiograph. 4 However, in its final publication and with the
objective of improving the recommended radiographic techniques, film speeds were changed
from 200—the speed used in most of the facilities where reference doses were measured—to
400, which made the correlation invalid!
Measurements of image quality and dose and their comparison with the standards should not
divert the Medical Physicist‘s search for the best combination of image quality and patient dose,
but foster it. Furthermore, not implementing image quality/dose standards may result in
regulatory authorities making compliance with diagnostic reference levels mandatory, regardless
of image quality. Regulators‘ zeal for dose reduction, and their potential to equate low doses in
194
diagnostic examinations with good quality, may be detrimental for patients undergoing a
radiological examination in a facility where the radiologist—because of his/her training—
requires low noise images for an accurate radiographic interpretation.
Rebuttal
There are many issues in Professor Goodenough‘s opening statement with which I fully agree. It
seems that we both abhor regulators, who ‗‗seem fond of imposing an upper limit on dose.‘‘ He
questions, ‗‗Why not a lower limit?‘‘ Indeed, why not? It all depends on how standards are
defined. The ‗‗International Basic Safety Standards for Protection against Ionizing Radiation and
for the Safety of Radiation Sources (BSS)‘‘ 3 states that ‗‗guidelines levels‘‘ are to ‗‗be used as
guidance by medical practitioners‘‘ and are to ‗‗be derived from data from wide scale quality
surveys.‘‘ It stresses that ‗‗corrective actions‘‘ are to ‗‗be taken as necessary if doses or activities
fall substantially below the guidance levels and the exposures do not provide useful diagnostic
information and do not yield the expected medical benefit to patients.‘‘ Also, ‗‗reviews‘‘ are to
‗‗be considered if doses or activities exceed the guidance levels as an input to ensuring optimized
protection of patients and maintaining appropriate levels of good practice.‘‘ However, ‗‗these
levels should not be regarded as a guide for ensuring optimum performance in all cases, as they
are appropriate only for typical adult patients and therefore, in applying the values in practice,
account should be taken of body size and age.‘‘ Characterized in these terms, who can argue that
these standards will not benefit patients? No wonder they were endorsed by the member states of
six international organizations!
I agree that the relationship between dose and image quality is a complex issue and that ‗‗more
than standards we need increased physician education of actual dose levels and the legitimate
dose trade off for image quality.‘‘ However, how is this education to be achieved if the
radiologists do not have the means of comparing and judging what they do? How many
radiologists are aware of the doses involved in certain procedures? How can they optimize
patient protection without dose/image quality relationships? Why can they not participate in their
formulation? Who says that standards are synonymous with regulations?
I disagree that ‗‗standards may set a ‗lower bar‘ of performance.‘‘ Why should they? Do we stop
when we reach a goal or do we use it as a stepping stone to further improvement? Alas, the
answer does not lie in physics, but in human nature with its inherent quest for perfection.
REFERENCES
1. ISO 9000 Series. ISO, Geneva (1994).

2. Conference of Radiation Control Program Directors, Inc. ‗‗Average Patient
Exposure/Dose Guides,‘‘ CRCPD, Frankfort (1992).
3. FAO/IAEA/ILO/NEA~OECD!/PAHO/WHO, ‗‗International Basic Safety Standards for
Protection Against Ionizing Radiation and for the Safety of Radiation Sources,‘‘ Safety Series
115. IAEA, Vienna (1996).
4. European Commission, ‗‗European Guidelines on Quality Criteria for Diagnostic
Radiographic Images,‘‘ EUR 16260 (1996).
195
4.6. Diagnostic ultrasound should be performed without upper

intensity limits
William D. O'Brien, Jr. and Douglas Miller
OVERVIEW
As with most diagnostic technologies, ultrasound imaging reflects a trade-off between image
resolution and energy absorption in tissue. With diagnostic ultrasound, current upper limits on
beam intensity have not been correlated with demonstrated harmful effects. Microcavitation has
been observed at intensities near these limits, but its biological significance is unknown. This
Point/Counterpoint explores whether upper intensity limits should be removed to permit
improvements in the quality of ultrasound images.
Arguing for the Proposition is William D. O'Brien, Jr., Ph.D. Dr. O'Brien is Professor of
Electrical and Computer Engineering and of Bioengineering, College of Engineering; Professor
of Bioengineering, College of Medicine; and Director of the Bioacoustics Research Laboratory at
the University of Illinois. Previously, he worked at the Bureau of Radiological Health (currently
the Center for Devices and Radiological Health) of the U.S. Food and Drug Administration. He is
a fellow of four professional societies; has served as president of the IEEE Ultrasonics,
Ferroelectrics, and Frequency Control Society and the American Institute of Ultrasound in
Medicine; and is Editor-in-Chief of the IEEE Transactions on Ultrasonics, Ferroelectrics, and
Frequency Control. His research interests involve the many areas of ultrasound–tissue
interaction, including spectroscopy, risk assessment, biological effects, tissue characterization,
dosimetry, and imaging for which he has published 215 papers.
Arguing against the Proposition is Douglas Miller, Ph.D. Dr. Miller is a Senior Research
Scientist at the University of Michigan Department of Radiology. He received a Ph.D. in Physics
from the University of Vermont in 1976, and worked at Battelle Pacific Northwest National
Laboratory on bioelectromagnetics and ultrasonic biophysics research before moving to
Michigan. Dr. Miller has served on ultrasound safety review groups of the American Institute of
Ultrasound in Medicine, World Federation of Ultrasound in Medicine and the National Council
on Radiation Protection and Measurements. Present NIH projects include research on the
bioeffects associated with contrast aided diagnostic ultrasound and on ultrasound enhanced
cancer gene therapy.
FOR THE PROPOSITION: William D. O'Brien, Jr., Ph.D.
Opening Statement
Regulatory control of diagnostic ultrasound equipment in the U.S. can be traced to passage of the
1976 Medical Device Amendments to the Food, Drug, and Cosmetic Act. When the FDA
initiated the regulation of diagnostic ultrasound equipment in its 1985 "510(k) premarket
notification," application-specific intensity limits were set that manufacturers could not exceed.
196
The 510(k)'s purpose was for the FDA to assess if a new device was "substantially equivalent," in
safety and effectiveness, to diagnostic ultrasound equipment on the market prior to 1976.
However, the intensity limits were not based on safety or effectiveness but rather on the
maximum intensity limits of diagnostic ultrasound equipment at the time when the Amendments
were enacted, in 1976; hence the term pre-amendments levels. To emphasize the FDA's date-
based regulatory approach, as opposed to safety and efficacy based, the American Institute of
Ultrasound in Medicine notified the FDA in mid-1986 that there existed prior to May 28, 1976 at
least two diagnostic ultrasound devices with intensity levels greater than the 1985 application-
specific intensity limits. In early 1987, the FDA updated their limits to the higher intensity levels.
Following widespread approval of the voluntary Output Display Standard (ODS) in early 1990,
the FDA essentially adopted the ODS for its regulatory guidelines. The ODS did not include
upper limits. Nevertheless, the FDA added application-nonspecific guideline upper limits that
were still based on the 1976 pre-amendments levels.
Problems with the date-based upper-limit regulatory approach include (1) a complicated set of
rules and procedures by which manufacturers verify to the FDA that their equipment is in
compliance, and the costs associated with these requirements; (2) a perception that these upper
limits are safe; (3) a demonstrated lack of attention to ODS education materials about the safety-
based biophysical indicators; (4) the exposure of patients at these upper limits for which there
may be safety concerns; (5) a limiting of future clinical benefits by preventing the development
of more advanced diagnostic ultrasound systems at higher levels; and, finally, (6) a recognition
that limiting diagnostic ultrasound capabilities may, in fact, be responsible for greater patient risk
due to either an inadequate diagnosis, or to the use of an additional diagnostic procedure for
which there is a defined risk.
The elimination of the upper-limit regulatory approach would have the following benefits: (1) a
less complicated set of rules and procedures by manufacturers, and at less cost; (2) the
elimination of the perception that there are safe limits; (3) more attention to the ODS education
materials; (4) more attention to the ODS-based biophysical indicators; (5) making available
research opportunities to develop advanced diagnostic procedures; and (6) providing the
diagnostic capability to obtain an adequate diagnosis if higher levels are required.
To apply rigid controls to ultrasound intensity without a proper scientific justification benefits no
one, particularly the patient. The physician is a professional trained to provide health care by
making informed benefit–risk judgements. The FDA's regulatory approach had denied the
physician the need to become informed about such benefit–risk issues, and for that we are all
worse off.
Rebuttal
The current government-mandated upper-intensity-limit regulatory approach has placed the risk
side of the risk–benefit decision on the FDA, not with physicians trained to make such decisions.
This is not how good medicine should be practiced. In an ideal world, the government would
protect us. The government consists of individuals like you and me, and none of us have the
knowledge or wisdom to know how to provide long-distance protection. Protection goes well
beyond making sure that diagnostic devices do not produce any bioeffects. Protection must
include on-the-scene decision making. In other words, a fundamental clinical issue is an accurate
and safe diagnosis of the patient. That is why physicians receive extensive training in the risk–
benefit decision making process.
197
My opponent argues that risk can be eliminated by limiting outputs to values below the threshold.
Obviously, this refers to ultrasound-induced risk. What about the risk associated with an
insufficient diagnostic quality image? What about the follow-on diagnostic procedure that might
have a significant hazard? We cannot view risk narrowly. It must be viewed in the broadest
sense, that of providing the best diagnosis of the patient.
My opponent also argues that limits provide a conceptual separation between diagnostic and
therapeutic uses of ultrasound, and that built-in safety limits free the operator from basic issues
such as safety, equipment operation, and complex dose calculations. Arbitrary boundaries
between types of equipment are artificial, but operator training is not. Users must have
appropriate training. The safety issue must be argued on what is best for the patient's health.
My opponent finally argues that without limits, manufacturers would be prone to engage in
competition for higher power for anticipated marginal increases in image quality. Even if these
were true, and this is unsupported, the marginal increase in image quality might make the
diagnostic difference for some patients. Is this not worth it? Should we not try it?
AGAINST THE PROPOSITION: Douglas Miller, Ph.D.
Opening Statement
The use of upper limits appears to be an ideal way to promote the safety of diagnostic ultrasound
examinations. Bioeffects of ultrasound occur by way of indirect mechanisms, such as heating or
acoustical cavitation, and appear to have identifiable thresholds. The risk of such an effect can be
eliminated by limiting outputs to values below the threshold. Of course, the exact forms and
values of upper limits on instrument output should be continually questioned and improved. The
present guideline limits1 are arbitrarily linked to the state of the art in 1976, rather than to
scientific principles.
However, through deliberation and consensus within the ultrasound community, the restrictive
limits originally applied to many examinations have been raised to the benefit of effective
diagnosis. In addition, on-screen readouts of thermal and mechanical exposure indices provided
on many recent systems are valuable for the identification and management of worse-case
conditions, for which some potential for bioeffects exists with current machines both from
heating (most likely of bone) and from mechanical effects (most likely on lung or intestine).
The use of well-chosen upper limit guidelines for diagnostic ultrasound also has several other
benefits. Limits provide a conceptual separation of diagnostic and therapeutic uses of ultrasound,
thus delineating regimes for patient safety on the one hand, and for effective treatment (e.g.,
physical therapy, surgery, or lithotripsy) on the other hand. Simple built-in safety limits free the
operator from complex dose calculations, detailed risk/benefit determinations or extensive safety
training for use of an instrument. Furthermore, authoritative upper limits engender a degree of
public confidence in the safety of examinations not possible without them.
Eliminating upper limit guidelines is a bad idea for several reasons. If this were done, instrument
manufacturers would be prone to engage in competition for higher power for anticipated
marginal increases in image quality (e.g., a doubling of intensity only provides an additional 6%
imaging depth, even for linear propagation2). Alternately, higher output power might be
198
substituted for expensive receiver amplifiers, thus increasing risk with no real patient benefit.
Most importantly, it seems likely that patient harm would eventually be encountered.
In summary, guideline upper limits for diagnostic ultrasound have many advantages, and even
the present imperfect limits have served manufacturers, physicians and patients well. Eliminating
limits, which would likely be of only marginal value for improving image quality, would also
eliminate the confidence of the medical community and the general public in the safety of
diagnostic ultrasound examinations. Diagnostic ultrasound should not be performed without
upper intensity limits.
Rebuttal
Dr. O'Brien has clearly stated the problems perceived to be associated with the existence of
guideline upper limits for diagnostic ultrasound. However, removing the present limits would not
solve the problems he has listed because (1) manufacturers would still have to follow
complicated procedures in order to provide the output display indices for FDA clearance; (2) the
present limits are justified, to some extent, by extensive research and through considerations of
bioeffects mechanisms; (3) the elimination of limits would do little to advance knowledge or
education about safety issues; (4) safety concerns related to present outputs would certainly not
be ameliorated by the removal of upper limits; (5) in fact, higher levels are not prohibited, and a
truly compelling new device requiring levels in excess of the guidelines can gain FDA approval
by demonstrating safety and efficacy; (6) greater advances in diagnostic ultrasound would be
obtained by improving training in imaging procedures, safety issues, and diagnostic
interpretation than by slight improvements in images at higher intensities.
The removal of guideline limits would permit manufacturers to design diagnostic ultrasound
instruments to engineering limits regardless of safety issues, even though the engineering
objective may not serve more general medical needs and desires. Physicians and sonographers
should focus on the art of sonography and diagnosis rather than on complex safety issues related
to the selection of an intensity for each examination. Many patients do not receive any
discernible benefit from a diagnostic test, and consequently expect to be protected from
unnecessary levels of risk. The present upper limit guidelines, though less than perfect, satisfy
these general needs and work well in the real world of incomplete safety information and
inadequate education. Manufacturers, sonographers, and patients all benefit from the framework
provided by the existing approval process. Guideline upper limits for diagnostic ultrasound
should be continually improved, but should not be removed.
REFERENCES
1. Information for manufacturers seeking clearance of diagnostic ultrasound systems and

transducers." CDRH, U. S. Food and Drug Administration, Rockville, MD, 1997.
2. NCRP, "Exposure criteria for medical diagnostic ultrasound: I. Criteria based on thermal
mechanisms." National Council on Radiation Protection and Measurement, Bethesda, MD, 1992.
199
CHAPTER 5
X-Ray Imaging
5.1. In the next decade automated computer analysis will be an
accepted sole method to separate "normal" from "abnormal"
radiological images
Kenneth R. Hoffman and Joel E. Gray

Reproduced from Medical Physics, Vol. 26, No. 1, pp. 1-4, January, 1999
OVERVIEW
Computer-assisted analysis of radiographs has proven effective in enhancing the detection of

lesions, especially in mammographic and chest images. Most articles describing this technology
have emphasized that automated computer analysis is adjunctive to, and not a substitute for,
review of films by a radiologist. Yet ROC curves for lesion detection with automated analysis
alone are comparable to those of radiologists alone. In comparing these techniques as a method
to separate ‗‗normal‘‘ from ‗‗abnormal‘‘ images, one should recognize that computer analysis
does not have to be perfect; it needs to be only as good as the typical radiologist. This issue of
Point/Counterpoint examines the future feasibility of using computer analysis alone as a way to
separate ‗‗normal‘‘ from ‗‗abnormal‘‘ images.
Arguing for the Proposition is Kenneth R. Hoffman. Dr. Hoffmann received his Ph.D. in solid
state physics from Brandeis University in 1984. He joined the Kurt Rossmann Labs at the
University of Chicago in 1984. He is currently an Associate Professor of Radiology at the
University of Chicago. He has been active in research in vascular imaging, 3D reconstruction
from biplane and single-plane images, computer-aided diagnosis (CAD) of lung diseases from
single projection images, and analysis of CT data sets with the goal of facilitating application of
CAD techniques to 3D data sets.
Arguing against the Proposition is Joel E. Gray. Dr. Gray received his Ph.D. in Radiological
Sciences from the University of Toronto in 1977. He joined the Department of Radiology, Mayo
Clinic where he was a Consultant and Professor in the Mayo Graduate School of Medicine until
1997. He introduced the concept of quality control to medical imaging in 1976 and published
Quality Control in Diagnostic Imaging, and coauthored the ACR Mammography QC Manual.
He has published over 130 papers in peer-reviewed journals on most aspects of diagnostic
medical physics. Dr. Gray is an independent consultant to healthcare organizations, industry, and
government in Medical Technology Management and Assessment, Medical Physics, and
Imaging Sciences.
200
For the proposition: Kenneth R. Hoffman
Opening Statement
Computer analysis is already being used to separate normal from abnormal images. The many
receiver operating characteristic (ROC) trials of various computer-aided diagnosis schemes rely
on computer analyses to specifically perform this task. Over the years, a number of these
analysis techniques have improved in accuracy, and, more importantly, in sensitivity and
specificity, that is, the area under the ROC curve (the classification performance) for these
analyses has increased. Some of these techniques are currently outperforming residents and
attendings, though not the experts, not yet. A number of investigators (and study section
members) see opportunities for improving these techniques further by using neural network
training and case-based reasoning because of their similarity to human ways of learning.
So far, most computer analyses perform a specific task, usually a binary task of evaluating
images for a specific disease. To date, no computer algorithm can take an image and evaluate
whether and which disease is present. For this evaluation, more than the image information alone
must be presented to the computer, as it is to the radiologist. Indeed, many computer analyses
have become more sophisticated with clinical information incorporated with image information
to improve further classification performance. Some preliminary results for differentiating
specific types of lung disease have appeared. We can expect analysis techniques to continue
improving. Cognizant of the history of computer analysis of radiographs, can we rationally say
that for some (and perhaps increasing numbers of) diseases, normals and abnormals will never be
separated by the computer alone? Will radiologists always defeat a ‗‗Big Blue‘‘ of radiology? It
seems that the question is not whether, but when, will computers be allowed to read radiographic
images in place of radiologists?
The answers to these questions depend on the criteria that are used to determine whether (when)
computers can be left alone to diagnose groups of patients or a particular patient. When would
you want a computer evaluating your radiograph as opposed to a nonradiologist, an attending
who reads those types of studies occasionally, or an expert? What are you willing to pay? For
yourself or as part of society?
At least in part, the issue here is ‗‗alone.‘‘ Does the computer have to replace the radiologist in
all aspects of diagnosis, i.e., be left alone, without the occasional intervention by expert
radiologists? Resident physicians do not function in this manner. Attendings refer to colleagues
with greater expertise in various areas. Radiologists receive certification (of expertise) for
various specialties. Should we not think about applying the same standard to the computer? Are
there (initially) simpler tasks which could be performed by the computer?
Could the computer be asked, for example, to ‗‗rule out pneumonia‘‘? When should computers
be allowed to make the decision to call in an expert, as an attending would do?
We need to answer these questions, and we will. It is a question of when.
Rebuttal
201
With the last sentence in his initial argument, Dr. Gray implies that it is no longer a question of
whether computer technology will be used to differentiate normals from abnormals, rather it is a
question of when to ‗‗bless‘‘ it.
Computer aided diagnosis (CAD), as a second opinion, has already been shown to improve
diagnostic accuracy. False positive rates continue to drop as techniques improve (for example, in
mammography), and studies indicate that false positives from CAD have not resulted in
increased call backs. When techniques outperform individual radiologists in detection and
classification ~as some currently do!, are we to deny patients the right to a more accurate
diagnosis? Are we willing to pay the additional emotional and financial costs when we do not
use the results of the (more accurate) computer analysis? Who will be willing to tell the patient
that the film has to be read by a radiologist, especially when that radiologist may be less accurate
than the computer?
Results indicate that with CAD diagnostic accuracy can be increased while reducing the time for
diagnosis, primarily by facilitating sorting out the ‗‗normals.‘‘ When ‗‗for sure normals‘‘ are
determined with greater accuracy than a given radiologist can achieve, will that radiologist not
want to use the computer and have more time to inspect the ‗‗abnormal‘‘ cases? Will we demand
that a radiologist read every film, even though the computer reads at least some subset of films
more accurately and quickly? The experience with Pap smear testing may provide some insight
and a precedent.
Use of the computer for diagnosis will demand increased involvement by the medical physics
community. This technology needs to be developed further, evaluated properly, and monitored
carefully. Medical physicists in particular have an obligation to patients to perform these duties
specifically because of their understanding of the variety of issues and their expertise in
addressing them.
Against the proposition: Joel E. Gray
Opening Statement
Computer assisted diagnosis (CAD) has been in the wings for many years. Hopefully it will stay
in the wings! Although several groups are working to ‗‗perfect‘‘ CAD, it is doubtful that it will
ever be widely adopted by radiologists. Would medical physicists accept software that would
make their jobs unnecessary? A dosimetrist with software could easily replace a medical
physicist in radiation oncology—but who would benefit from this? The administration may save
a few dollars on the medical physicist‘s salary. However, elimination of the professional input
from highly skilled medical staff (be they radiologists or medical physicists) is not in the best
interest of the patient and quality healthcare.
Proponents claim that CAD will improve diagnostic accuracy, i.e., lesions will not be overlooked
by the radiologist. However, CAD systems are not perfect and introduce false positives. These
false positives increase the time the radiologist must spend on a particular case. They also
increase the number of biopsies or other diagnostic procedures, thereby increasing the cost of
medical care. Health economists tell us that there are already too many biopsies in
mammography, and that we should be doing everything possible to reduce the biopsy rate.
What is the advantage to a diagnostic radiologist to have cases separated into ‗‗normal‘‘ or
‗‗abnormal‘‘? Will the typical radiologist trust the computer and sign-off on ‗‗normal‘‘ films
202
without reading them? Who gets sued for malpractice if the patient with a ‗‗normal‘‘
examination (which was not read by a radiologist) is found to have malignant disease six months
later, and the lesion was present in retrospect in the ‗‗normal‘‘ film? If the radiologist trusted the
computer and signed off on the film, then the radiologist will be spending a lot of time with
attorneys.
And what about ‗‗abnormal‘‘ films? The radiologist will, by necessity, spend a lot more time
looking at these ‗‗abnormal‘‘ films, especially when the computer indicates disease in a region
that appears normal to the radiologist. ROC curves clearly demonstrate that we can reach
perfection (100% true positive rate) if we are willing to tolerate an increase in the number of
false positive readings. However, each and every false positive comes at an additional cost, both
emotional and financial. The emotional cost is borne by the patient who is told there is a
suspicious area in her breast and a biopsy is recommended. From the time these fateful words are
spoken until the biopsy results prove that there is no disease, the patient will be in such
emotional turmoil that words cannot possibly describe her feelings. The financial cost will be
borne by the insurance company or healthcare system, which ultimately means the patient. All of
us will pay increased premiums to cover the cost of more biopsies or additional examinations to
assure the radiologist that disease is not present.
So what are the reasons for wanting to adopt this technology? Just because it is computer based
does this mean it is better for patient care? Is technology driving our healthcare decisions, or do
we control the technology that will be used? The proponents of full breast digital mammography
are required by the FDA to demonstrate that the diagnostic capability of this ‗‗new modality‘‘ is
at least as good as screen- film mammography. As a result, several controlled clinical trials are
being carried out by manufacturers to compare screen-film and digital mammography. Why has
CAD not undergone similar scrutiny? Where are the clinical trials of several thousand patients
that indicate the efficacy of CAD, the added cost in terms of the radiologist‘s time, and the
increased number of biopsies or other additional diagnostic examinations?
We should not accept technology for the sake of technology. CAD may offer benefits in the
future, but it is unclear at this time exactly what those benefits are. It is time for extensive
clinical trials before the FDA blesses this technology for use in the diagnostic armamentarium.
Rebuttal
While Dr. Hoffman and I agree that computer assisted diagnosis (CAD) is not ready for prime
time (not ready for clinical application), our views diverge on most of the other issues associated
with this topic. Software available today can sort unknown films into ‗‗normal‘‘ and
‗‗abnormal‘‘ for limited disease states. However, the radiologist is normally faced with the
potential for any of a number of diseases and in some cases, multiple diseases, something which
cannot be addressed by today‘s software. Most importantly, how can software handle the myriad
of potential diseases or conditions often discovered and referred to as ‗‗incidental findings?‘‘
Can software locate a fractured bone when it is looking for pulmonary nodules; a soft tissue
tumor when it is looking for a fractured femur; or tuberculosis when ruling out pneumonia?
These are the tasks which separate the non-radiologist physician from the radiologist when
reading films, or separate the radiologist from the radiology-expert.
Dr. Hoffman asks if you are willing to pay for a non-radiologist or an attending to read your
films, or if you want the expert to read them. This is not really the appropriate question! The real
203
question is ‗‗Are you willing to pay the price of a poor or missed diagnosis when the expert did
not read your films?‘‘ This price could be additional tests, unnecessary surgery, or the possibility
of a serious condition not being diagnosed in a timely manner leading to complications or worse.
When CAD reaches the performance level of the expert this will not be an issue. In fact, when
CAD reaches the expert level then all films should be ‗‗read‘‘ by the computer. Unfortunately,
this is not the time—and that time will not come for at least another ten years!
204
5.2. In x-ray computed tomography, technique factors should be

selected appropriate to patient size
Walter Huda and Stewart C. Bushong

OVERVIEW
A USA Today article in mid-January claimed that 1500 deaths will occur annually as a
consequence of CT exams to the heads and abdomens of children. The article states that most of
the deaths could be avoided by "calibrating" the scanners (i.e., using technique factors)
appropriate for children rather than adults. A few physicists had advocated the adjustment of CT
technique factors to patient size well before appearance of the USA Today article. Others feel that
the article is editorial hyperbole, and that there is no compelling reason to alter technique factors.
This controversy is the topic of this month's Point/Counterpoint.
Arguing for the Proposition is Walter Huda, Ph.D. Dr. Huda obtained his Ph.D. in medical
physics at the Royal Postgraduate Medical School (Hammersmith Hospital) at the University of
London. Dr. Huda worked for five years at Amersham International, a company specializing in
radioactive products. In 1982 he moved to North America, and worked in medical physics at the
Manitoba Cancer Treatment and Research Foundation (Winnipeg) and the University of Florida
(Gainesville). Dr. Huda is currently Professor of Radiology at SUNY Upstate Medical University
(Syracuse), and Director of Radiological Physics. His research interests are in medical imaging
and radiation dosimetry, and he has published over 150 papers.
Arguing against the Proposition is Stewart Bushong Sc.D. Dr. Bushong obtained his doctorate at
the University of Pittsburgh and then joined the faculty of Baylor College of Medicine where he
has enjoyed an exceptionally productive 35 years. During that time he has published over 125
scientific papers and 30 books. Radiologic Science for Technologists, now in its 7th edition, is
the standard text for such training programs. He has always been active in the affairs of the
AAPM and ACMP.
FOR THE PROPOSITION: Walter Huda, Ph.D.
Opening Statement
X-ray CT scanners are digital imaging systems that offer users a wide range of latitude in terms
of technique factors (i.e., mAs and kVp). Choice of technique factors affects the image contrast
to noise ratio (CNR) and patient dose, both of which are dependent on the size of the patient.
Increasing the patient weight from 10 to 120 kg for abdominal CT scanning reduces the
transmission of x-ray intensity by about a factor of 100. 1 Maintaining the same radiographic
technique factors results in the CNR being determined by the size of the patient!2 This state of
affairs is illogical, since CNR should be determined by the imaging task at hand. Maintaining the
same technique factors for infants substantially increases image CNR.1 There are no data
205
whatsoever that this procedure improves diagnostic performance. On the contrary, increasing
evidence indicates that CT technique factors could be reduced, with no loss of image quality.3
In large US hospitals, CT scanning accounts for more than 10% of diagnostic radiology
examinations and two thirds of the radiation dose. 4 The lifetime cancer mortality risk for a one-
year-old patient attributable to the radiation exposure from an abdominal CT is estimated to be
0.18%.5 Individual and collective CT doses are substantial, and need to be taken seriously.
Although the issue of radiation risk associated with low doses is controversial, it is difficult to
argue against the as low as reasonably achievable dictum recommended by the International
Commission on Radiological Protection. Given that a CT examination is justified by the benefits
to the patient, there is no need to use more radiation than is required to obtain a diagnostic image.
There is little doubt that diagnostic CT scans of infants and children could be acquired using
much lower technique factors,1 with substantial savings in patient dose. Additional benefits
include longer x-ray tube lifetimes and longer scan lengths in helical scanning. Taking into
account the size of the patient is simple to achieve, 6 eliminates unnecessary radiation to patients,
and carries no significant penalty. Modifying CT techniques to account for patient size is clearly
the "no brainer" of a recent editorial in the AJR.7
Rebuttal
There appear to be many areas of agreement between Dr. Bushong and myself. Radiographic
technique factors should be adjusted to take into account patient size and to eliminate
unnecessary radiation exposure. There are large uncertainties in current radiation risk estimates,
and it is theoretically possible for "benefits" to exist at low radiation levels (hormesis). There are
substantial benefits to be gained from radiographic imaging procedures such as CT, but these
should always be weighed against estimated radiation risks.
At organ doses of tens of mGy associated with CT examinations, there are no definitive scientific
data related to (any) radiation risks. There is also insufficient knowledge of radiation-induced
carcinogenesis to make theoretical predictions. Accordingly, the scientific community cannot
quantify radiation risks from CT examinations. Furthermore, determining what level of risk is
"acceptable" is not a scientific issue.8,9 Society as a whole, through the normal political process,
must decide how to deal with current uncertainties in the scientific understanding of radiation
risks. My (political) preference is to agree with the ICRP philosophy which requires that
exposures be justified by the benefit received by the patient, and elimination of all unnecessary
radiation.
Historically, medical physicists working in diagnostic imaging have paid considerable attention
to equipment specification, commissioning of imaging systems, and ongoing QC programs.
Imaging equipment is becoming increasingly complex, and patient-imaging protocols would
benefit from explicit input from medical physicists. As imaging and radiation experts, medical
physicists should help ensure that patients are examined in a manner that ensures optimal image
quality. This requires the signal-to-noise ratio to be sufficient for the diagnostic task at hand, and
that patients receive no more exposure than is required for diagnosis.
AGAINST THE PROPOSITION: Stewart Bushong, Ph.D.
Opening Statement
206
At 7:30 a.m. on Monday January 22, 2001 I answered the phone and it was Robert Malone, M.D.
By 9:00 a.m. that morning I had received three more phone calls. Inquiries by radiologists and
patients haven't stopped yet. Headlines on the front page of that day's USA Today shrieked "CT
scans in children linked to cancer later." "What does this mean?" "What do I tell my patients?"
"Who is David Brenner?" The answers are quite simple but the damage done to CT imaging, and
maybe to x-ray imaging in general, will be hard to correct.
I agree with Lee Rogers7 that this is "a no brainer." Regardless of the x-ray examination,
technique factors should be adjusted to body size or body part. This is basic training for
radiologic technologists. The paper by Donnelly et al.6 is good medicine for facilities that do not
now practice technique optimization. The paper by Paterson et al.10 should not have been
published. Check the Results section "(chest and abdomen) medical CT examinations (n =
58)....and children (n = 32)" "in one (11%) of these examinations "
Based on a study population of 90, conclusions are suggested for the US population which
allowed USA Today to report "about 1.6 million children in the USA get CT scans and about
1500 of those will die later in life " This is not the experience at the CT imaging facilities of
the Texas Medical Center.
Brenner's paper5 and the two companion papers6,10 in the February issue of the American Journal
of Radiology conveyed the correct message. If CT technique is not adjusted to patient size, some
patients, especially children, will be exposed to unnecessary radiation dose. Fair enough! A good
message for those few facilities that do not presently reduce technique for pediatric patients.
However, Brenner performs calculations suggesting that the alleged unnecessarily high dose in
pediatric examinations causes 1500 excess cancers in children. This is the message that was
picked up by USA Today.
If a single pediatric CT examination is refused by a parent worried about the consequence of a

CT radiation exposure, far greater harm could be the result.
I wish to thank David Brenner for the reminder about dose efficiency. However, why couldn't he,
or someone, title the paper "Estimated Benefits " rather than "Estimated Risks "? The
flawed exercise regarding risk may be academically satisfying but is having the negative effect of
replacing patient care with patient scare. The application of BEIR WT coefficients without
acknowledging the current raging debate over the linear nonthreshold dose response
relationship,11,12,13 and the postulated existence of radiation hormesis, 14 results in a great
disservice to our patients and our colleagues. It is a one-sided statistical exercise that should have
been balanced with an outcome analysis of the benefit of pediatric CT examinations.
But then there would have been no headlines.
I could critically respond to Brenner's simplistic approach to radiation risk. His assessment is one
of an attributable risk. It is not a real risk. Much damage has been done by such exercise of
academic privilege. If Brenner had only stayed the course of his main conclusion—adjust dose
by CT technique to body size or part, as we do consistently for radiography.
Rebuttal
207
It is hard to argue with the stated proposition or with Dr. Huda's opening statement. Technique
optimization is a principle component of all radiography training programs. Radiologic
technologists certified in CT by the ARRT are expected to know technique optimization in CT.
Indeed, most CT operating consoles have pediatric techniques as part of the selection protocols.
I fear that the fuss generated by the AJR articles will lead to another layer of government
radiation regulations. I hope the ACR recognizes this possibility as it develops its CT
accreditation program.
Look what happened with mammography. First voluntary ACR accreditation, then mandated
federal regulations—MQSA. Clearly MQSA has been a success, but at what a cost—registration,
medical physics acceptance testing and continual performance evaluation, followed by the
inspection process.
So we are detecting more breast cancer earlier and making a dent on breast cancer mortality. The
question is, how much of this benefit can be attributed to MQSA? I suspect very little because
radiology is self driven to excellence in every area.
I have never seen an outcomes analysis of mammography relative to MQSA and I bet I never see
an outcomes analysis for pediatric CT regardless of radiation dose.
On the other hand, as I have preached to medical physicists may times radiation regulations
are good for us. They should be developed with four C's in mind: The more complex, the better;
the more confusing, the better; the more convoluted, the better; and, the more costly, the better.
MQSA could be renamed MPSA—Medical Physicist Security Act!
REFERENCES
1. W. Huda, E. M. Scalzetti, and G. Levin, "Technique factors and image quality as

functions of patient weight at abdominal CT," Radiology 217, 430–435 (2000).
2. J. F. Ende, W. Huda, P. R. Ros, and A. L. Litwiller, "Image mottle in abdominal CT,"
Invest. Radiol. 34, 282–286 (1999).
3. J. G. Ravenel, E. M. Scalzetti, W. Huda, and W. Garrisi, "Radiation exposure and image
quality in chest CT," AJR, Am. J. Roentgenol. 177, 279–284 (2001).
4. A. F. Mettler, P. W. Wiest, J. A. Locken, and C. A. Kelsey, "CT scanning: Patterns of use
and dose," J. Radiol. Prot. 20, 353–359 (2000).
5. D. J. Brenner, C. D. Elliston, E. J. Hall, and W. E. Berdon, "Estimated risks of radiation-
induced fatal cancer from pediatric CT," AJR, Am. J. Roentgenol. 176, 289–296 (2001).
6. L. F. Donnelly et al., "Minimizing radiation dose for pediatric body applications of
single-detector helical CT: Strategies at a large children's hospital," AJR, Am. J. Roentgenol.
176, 303–306 (2001).
7. L. F. Rogers, "Taking care of children: Check out the parameters used for helical CT,"
AJR, Am. J. Roentgenol. 176, 287 (2001).
8. C. G. Brunk, L. Haworth, and B. Lee, Value Assumptions in Risk Assessment (Wilfred
Laurier University Press, 1991).
9. W. Huda, "What do radiation experts really think are reasonable dose limits for
occupational exposure?," RSO Magazine Nov/Dec, 22–24 (1998).
10. A. Paterson, "Helical CT of the body: Are settings adjusted for pediatric patients?," AJR,
Am. J. Roentgenol. 176, 297–301 (2001).
208
11. J. Ahlquist, "To LN-T or not to LN-T?—That is not the question," HP Newsletter 27(b),
6–9 (1999).
12. K. L. Mossman, "The linear nonthreshold debate: Where do we go from here?," Med.
Phys. 25, 279–284 (1998).
13. K. J. Arrow et al., "Is there a role for benefit-cost analysis in environmental, health and
safety regulation?," Science 272, 221–222 (1996).
14. E. J. Calabrese, "Radiation hormesis: Origins, history, scientific foundations," Belle
Newsletter 8(2), 1–66 (1999). first citation in article
209
5.3. Next-generation x-ray CT units will provide <500 msec images

with 3D resolution comparable to today’s projection radiography
Mahadevappa Mahesh and Dianna Cody

OVERVIEW
Some physicist savants believe that spiral scanning is just the beginning of a new generation of
CT units, and x-ray tubes with greater output and x-ray detectors with faster response will
provide images in the not-too-distant future that meet the specifications stated in the proposition.
Other thoughtful physicists consider this belief to be wishful thinking, not only because of
technology limitations, but also because it reflects a technology imperative that is unrealistic in
today's cost-conscious healthcare environment. This difference in opinion is the
Point/Counterpoint topic for this month's issue.
Arguing for the Proposition is Mahadevappa Mahesh, Ph.D. Dr. Mahesh is the Chief Physicist at
Johns Hopkins Hospital and is on the faculty at Johns Hopkins University School of Medicine,
Baltimore. Dr Mahesh obtained his Ph.D. in Medical Physics from the Medical College of
Wisconsin. He is past president of the Mid-Atlantic Chapter of AAPM, and serves on several
AAPM committees. Dr Mahesh is currently the Curator for the Partners in Physics program. He
is certified in Diagnostic Radiologic Physics from the American Board of Radiology, and is a
member of the Radiation Control Advisory Board for the State of Maryland. He was the director
for the continuing education course on CT and for the symposium on CT dose reduction at the
2002 AAPM annual meeting.
Arguing against the Proposition is Dianna D. Cody, Ph.D. Dr. Cody trained as a bioengineer at
the University of Michigan and attained board certification in Medical Physics (Diagnostic)
while on staff at Henry Ford Hospital in Detroit, Michigan. She spent 13 years at Henry Ford,
and in 2000 she joined The University of Texas M.D. Anderson Cancer Center in Houston,
Texas as an Associate Professor. She is Co-Director of the Image Processing Laboratory, and
provides CT physics support for 11 clinical scanners, 2 research scanners, and 4 scanners housed
in therapy. She is also involved in small animal CT imaging research, CT medical physics
graduate education, and CT related workshops for practicing medical physicists.
FOR THE PROPOSITION: Mahadevappa Mahesh, Ph.D.
Opening Statement
CT technology has evolved considerably over the past 30 years. Evolution of the technology can
be viewed as a progression towards optimizing the information content of CT images and
extending the method to three dimensions, i.e., to achieve isotropic resolution—the holy grail of
medical imaging.1 I believe that the technology is poised for further breakthroughs with scanners
capable of rotation speeds less than 500 msec, while providing very high resolution images as
210
stated in the proposition. In defense of the proposition, scanners reaching the marketplace can
already operate at 400 msec scan speeds with very high spatial resolution.2
Before spiral CT, critics thought that CT technology had reached a plateau and further
developments would not be cost effective. Moreover, true 3D reconstructions were thought to be
impractical in body regions subject to physiologic motion. However, technological advances
such as slip-ring gantries, high power x-ray tubes and interpolation algorithms produced a true
technological revolution.1 With early single row spiral CT scanners, 3D volume acquisitions
within a single breath-hold became possible. Still, resolutions approaching isotropic could be
attained only over very small volumes with very high tube loads.3 During the latter part of the
past decade, considerable progress has been achieved with introduction of multiple row detector
array CT (MDCT) (2-sections and 4-sections per rotation). Now, with introduction of 16+ section
scanners, it is possible to scan at sub-millimeter spatial resolution with high temporal resolution.4
The next generation of CT scanners will provide scan speeds of less than 500 msec per scan
rotation, aiming towards 100 msec scan speeds. The latter scan speed is ideal for cardiac imaging
without physiologic motion to yield very high spatial and temporal resolutions. Fabrication of
thin detectors (0.5–0.75 mm) with very fast response times (e.g., ceramic scintillators) will
enable acquisition of images of thin sections with high spatial resolution in all directions. In
addition, development of cone beam reconstruction algorithms, now capable of achieving 8 to 16
thin sections per rotation, will generate even more sections (32–40) per rotation.5,6 With modern
reconstruction algorithms it is possible to achieve very high temporal resolution on the order of
100–200 msec, enabling new protocols like CT angiography.7 Also, the number of detector rows
in the z-direction will increase from 16 to 40 to 64 and beyond. Already there is progress towards
fabricating even greater rows of detector elements (256 rows) to permit a larger volume
acquisition in a very short time. The arrival of flat panel detectors, development of sophisticated
cone beam reconstruction algorithms,7 and utilization of high speed computers to handle
voluminous scan data, are not very far in the future. These advances will be accompanied by
light-weight x-ray tubes with minimal moving parts, detectors with faster response, and even
faster computers. Rising health care costs might impede further spending on new technologies.
As seen in the past, however, the initial cost of new technologies does not tend to hinder the
ultimate progress of medical technology.
As stated in the proposition, new generation CT scanners will make high resolution, 3D CT
images a reality in the near future. These scanners will pave the way to newer clinical
applications, further widening the application of CT in medicine worldwide.
Rebuttal
Dr. Cody and I agree on the stated proposition; our differences are in the "time and challenges" in
achieving it. I agree that the next generation CT (it all depends on how "next generation" is
defined) may not be able to provide 3D images of quality comparable to projection radiography.
However, I respectfully disagree on some of the reasons she lists in her opening statement.
The goal of CT has always been to provide images of spatial resolution similar to projection
radiography, but in three dimensions. It is true that the cost of CT is greater than projection
radiography, but so too is the information it provides. The increased information from
multidetector CT has led to additional clinical applications, including several in cardiology. With
increasing utilization, the cost of the scanner and its maintenance shrink in comparison with the
revenue the scanner generates.
211
The increasing workload to process 3D data sets is certainly due to the data explosion of
multidetector CT. Novel methods to manage large data sets and workstations are being
developed. When cross-sectional CT was introduced, a relatively short time was needed for
clinicians to adjust to reading CT images. Similarly, clinicians will learn relatively quickly how
to handle 3D images. Yes, the initial costs of presenting and managing large data sets are higher,
but the effectiveness of 3D images on diagnosis will overcome this cost over time. At our
institution, the chief of surgery insists on seeing 3D images on his patients before performing
surgery. I believe it is only a matter of time before every clinician demands the same.
I agree that the radiation dose in CT is very important. Even though the clinical content of most
CT exams outweighs the radiation risks, one should carefully compare the risks and benefits
before implementing newer applications. We in the medical physics community have an
obligation to help our clinical colleagues perform better risk—benefit analyses for all new CT
protocols. Increasing scrutiny on this issue has led manufacturers to reduce patient dose. In the
near future, true automatic exposure-control CT systems may become available. It is always
essential to consider the appropriateness of every CT examination.
Shorter scan time and higher spatial resolution are critical to image quality, especially for rapidly
moving organs such as the heart. The current emphasis in CT technology is targeted primarily at
imaging cardiac anatomy and function. In closing, I wish to emphasize that even though the
challenges are many, they will be met and we will soon see very fast CT scanners capable of
producing 3D images with resolution comparable to projection radiography.
AGAINST THE PROPOSITION: Dianna Cody, Ph.D.
Opening Statement
Arguing against this proposition is a bit like stating that the sun won't rise tomorrow. Ultimately,
3D imaging will provide results comparable to today's projection radiography. However, the
challenges to this destiny are many and severe, and 3D imaging will not emerge with next-
generation CT units. Instead, this evolution will require several iterations in its progression.
The major challenges to the progression of events are:
Cost: Increasing CT equipment utilization will shorten its lifespan and increase its maintenance
cost. CT scans cost at least one order of magnitude more than projection radiographs. The current
relative value unit (RVU) assessments for a single view P/A chest and a no-contrast routine chest
CT exam are different by only about a factor of 6, due in part to the level of complexity of the
examination.8 Hardware and software needed to form 3D views would be required, along with
the infrastructure to produce 3D images on a routine basis.9 These requirements would further
increase the relative cost of producing a 3D study.
Radiation Dose: The radiation dose delivered by CT is currently under scrutiny.10 The difference
in skin dose between a routine P/A chest radiograph and a routine non-contrast chest CT is about
two orders of magnitude.11,12 Even if we could produce 3D CT data sets that are comparable in
resolution to radiographs today at this dose level, the question remains: Should we?
Workload: The time required to process a 3D data set is not trivial. Unless the data set
preparation can be done accurately and automatically, radiologists will probably insist on making
212
manual adjustments to the resulting display. They will also expect the original CT images to be
readily available for reference. The concept of interpreting only the 3D display in place of
original CT images may be attractive, but it is unrealistic.
Scan time and resolution: Current CT scanners provide CT images in under 500 msec. This time
limit constrains the 3D view to a short span (around 2 cm). A combination of faster gantry speed
and wider x-ray tube collimation may further increase CT scan speed. Current multi-row system
specifications claim 400–500 µm isotropic high-contrast resolution. This does not compare well
to projection radiography available with resolution of 30–200 µm, depending on the application.
Cross-sectional views do not include overlying anatomy and may not require as much resolution
to detect abnormalities. Acquiring more CT data faster, however, may require tradeoffs that
diminish low-contrast resolution.
There are other issues that should be considered. What about the cost of following up additional
false-positive findings that would emerge from the CT images? And will the 3D data set, or just
the 2D views of the 3D object, be saved to the picture archiving system (PACS)?
Although we are surely moving in the right direction, I believe that we are far from replacing
routine projection radiography with 3D CT imaging. Nevertheless, applications of 3D CT
imaging will likely continue to expand for specific diagnostic tasks.
Rebuttal
Dr. Mahesh has provided a concise summary of the evolution of CT technology, and has shared
his vision of future CT innovations. There is very little in his opening statement to disagree with.
I would like to point out, however, that Dr. Mahesh has limited his discussion of cost to new
technology development (buried in the CT scanner purchase price). The additional burden on
radiologists, as well as the cost of a technologist to generate 3D views, must also be considered
as an incremental expense over the price of obtaining and interpreting plane films. The need for a
3D workstation, an administrative infrastructure to manage the image processing activity, and
additional PACS capacity are also nontrivial costs.
Lastly, the increase in radiation dose delivered by CT compared with projection radiography
cannot be ignored. Currently in the U.S., concern about radiation exposure from CT exams is
focused primarily on children. This concern will eventually extend to adults as well.
Before 3D CT replaces projection radiography, the benefits will need to clearly offset the
drawbacks. This will need to be proven by carefully designed and conducted clinical research
studies.
REFERENCES
1. M. Mahesh, "Search for isotropic resolution in CT from conventional through multiple-

row detector," Radiographics 22, 949–962 (2002).
2. T. Flohr, K. Stierstorfer, H. Bruder, J. Simon, and S. Schaller, "New technical
developments in multislice CT—Part 1: Approaching isotropic resolution with sub-millimeter
16-slice scanning," Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 174, 839–845 (2002).
3. W. A. Kalender, "Thin-section three-dimensional spiral CT: is isotropic imaging
possible?," Radiology 197, 578–580 (1995).
213
4. K. Taguchi and H. Anno, "High temporal resolution for multislice helical computed
tomography," Med. Phys. 27, 861–872 (2000).
5. K. Stierstorfer, T. Flohr, and H. Bruder, "Segmented multiple plane reconstruction: a
novel approximate reconstruction scheme for multi-slice spiral CT," Phys. Med. Biol. 47, 2571–
2581 (2002).
6. T. Kohler, R. Proksa, C. Bontus, M. Grass, and J. Timmer, "Artifact analysis of
approximate helical cone-beam CT reconstruction algorithms," Med. Phys. 29, 51–64 (2002).
7. K. Nieman, F. Cademartiri, P. A. Lemos, R. Raaijmakers, P. M. Pattynama, and P. J. de
Feyter, "Reliable noninvasive coronary angiography with fast submillimeter multislice spiral
computed tomography," Circulation 106, 2051–2054 (2002).
8. 2002 National Physician Fee Schedule Relative Value File, Center for
Medicare/Medicaid Services.
9. L. Logan, "Seeing the future in three dimensions," Radiol. Technol. 72, 483–487 (2001).
10. National Council on Radiation Protection and Measurements, Symposium on Computed
Tomography Patient Dose, Crystal City, VA, November 2002.
11. S. C. Bushong, Radiologic Science for Technologists: Physics, Biology and Protection
(C. V. Mosby, St. Louis, 1980), p. 424.
12. L. H. Rothenberg and K. S. Pentlow, "CT Dosimetry and Radiation Safety," in Medical
CT and Ultrasound, Current Technology and Applications, edited by L. W. Goldman and J. N.
Fowlkes (Advanced Medical Publishing, Madison, 1995), p. 538.
214
5.4. Free-standing, self referral centers for whole-body CT screening

should be closed, or at least restricted to at-risk patients
Richard Morin and Michael Broadbent

OVERVIEW
Direct-to-consumer marketing of whole-body CT screening is occurring today in many

metropolitan areas. Free-standing CT screening centers are being sited in shopping malls and
other high-density public areas, and these centers are being advertised in the electronic and print
media. Some physicists believe that the benefit:risk and benefit:cost ratios are too low to justify
these centers, and they should be closed, or at least restricted to at-risk, physician-referred
patients. Others feel that free-standing centers simply reflect a free-market economy, and that the
principle of caveat emptor should prevail. Free-standing CT screening centers are the topic of
this month's Point/Counterpoint.
Arguing for the Proposition is Richard Morin, Ph.D. Dr. Morin is the Brooks-Hollern Professor,
Mayo Medical School. He received his Ph.D. from the University of Oklahoma in Medical
Physics. His dissertation concerned the use of Monte Carlo Simulation and Pattern Recognition
for Artifact Removal in Computed Tomography. He was Director of Physics in the Department
of Radiology of the University of Minnesota before joining Mayo Clinic in 1987. He is a Fellow
of the ACR and a Diplomate of the ABR. Dr. Morin is a former President and Chair of the Board
of the AAPM, and is currently a member of the ACR Board of Chancellors, chairing the
Commission on Medical Physics.
Arguing against the Proposition is Michael Broadbent, Ph.D. Dr. Broadbent has been a
consulting radiological physicist in the Midwest for 26 years. He received his Masters and Ph.D.
degrees from UCLA, and is certified in radiological physics by the American Board of
Radiology. He has been active in the RSNA/AAPM Physics Tutorial Program, the Illinois
Radiological Society's Mammographic Accreditation Program, and the American College of
Radiology's Mammographic Accreditation Program. He is on the Medical Physics staff at Rush
University and assists in the instruction of Radiology residents.
FOR THE PROPOSITION: Richard Morin, Ph.D.
Opening Statement
This is America. Individuals are allowed to open any kind of business they wish as long as no
law, rule, or regulation is trespassed. How does this freedom apply to self-referral centers for
whole-body CT screening? First, such sites are banned by regulation in many states.1,2 That fact
notwithstanding, this controversial issue deserves thoughtful reflection. Medical screening
generally implies the application of an inexpensive low-risk test to an entire or large segment of
an asymptomatic population.3,4 Clearly, CT is not inexpensive, and scanning the entire or even a
substantial segment of the population is not practical. Hence, even the terminology "CT
215
screening" is misleading, or at best controversial. So why have such facilities and procedures
emerged? The answer is clear: money! Oft-heard responses to this statement are: "Live and let
live" and "No harm, no foul." But harm can be done. The results of false positive and false
negative studies may cause severe anxiety, or a false sense of security in the individual. In either
case, the best interest of the patient has not been served.
The follow-up of false positive findings in CT screening will cost the healthcare system millions
of dollars. Screening 100 000 individuals for a disease which has a 1% prevalence, and for which
CT has a 95% sensitivity and specificity, will yield about 5000 false positives. 5 Even if the
prevalence is increased by an order of magnitude to 10%, the number of false positives will be
4500. The emotional and financial cost of false positives to both the patient and society is
substantial. In addition, while the case may be made that the radiation risk is low for a 1–10 mSv
examination, such widespread exposure for limited expected benefit is contrary to the
fundamental principles of radiation safety.
These observations arise from a very fundamental issue. There is no scientific evidence that
whole-body "CT screening" is worthwhile.6 In fact, the major societies and agencies have issued
public statements against the use of whole-body "CT screening." 7,8,9,10 The absence of medical
evidence supporting the benefits of "CT screening" often brings this discussion to a boutique
level of individual antidotes and aggressive marketing. While an understanding of this issue may
not be present yet in the general public, readers of this journal certainly understand these
principles. It is precisely in this light that the truth is revealed. Whole-body "CT screening" is a
business, not good health care.
We end as we began. This is America. Individuals are free to open any kind of business they
wish. Pretending that the business of "CT screening" is good health care is disingenuous at best,
and scandalous at worst.
Rebuttal
My colleague has raised several interesting points. I concur with the assembly of statistics
regarding the incidence and effects of cancer in the USA.
Unfortunately, the point raised in reference to the Mayo Clinic study11 overlooked a few
important details. First, this study dealt with lung CT screening of a population of 50+ year olds,
each with 20 pack-years of cigarette smoking history. It did not consider whole-body CT
"screening." Second, over 2200 nodules were followed and 25 were malignant. The authors
conceded their concern about the "very high false-positive rate" 11 in this highly pre-selected
population, and called for a randomized prospective controlled study. Third, the lead author later
stated12 that it is unclear if CT scanning is an effective screening test,13 and called for a
randomized controlled trial to answer this question.
Dr. Broadbent refers to possible radiation hormesis benefits from whole-body CT "screening." If
these benefits exist, then whole-body CT "screening" should be marketed or prescribed on this
basis.
Informed individuals are free to accept risks in their daily activities. However, there is serious
doubt14 about the degree of "informed consent" in screening procedures. In this setting, the issues
of morality and risks discussed by Dr. Broadbent are questionable.
216
The analogy of whole-body CT "screening" to screening mammography is flawed. Screening

mammography is the most heavily regulated imaging procedure in the USA. Whole-body CT
"screening" is not even close. My counterpart has not really addressed the counterpoint. Whole
body "CT screening" may be good business but it is not good healthcare.
AGAINST THE PROPOSITION: Michael Broadbent, Ph.D.
Opening Statement
It is estimated by the American Cancer Society that 1 334 100 new cases of cancer will be
diagnosed in the United State during 2003. The Society also estimates a death rate from cancer of
206 per 100 000 persons. The leading causes of death in 2000 were heart disease, cancer and
cerebrovascular disease, with a total of 1 431 512 deaths (56 percent of all deaths).15 The use of
electron beam and conventional CT scans of the total body for lung cancer, colon cancer, and
coronary artery disease could potentially lead to the early detection and cure of many of these
patients.
Observational data, particularly for lung cancer, suggest that screening could have a major
benefit in outcomes despite the dearth of randomized data. "We know that if people with stage 1
lung cancer are not treated, about 80 percent of them will be dead in five years," says Dr. Brant-
Zawadzki, professor of radiology at Stanford University School of Medicine. "Compare this to
people with stage 1 lung cancer who undergo surgery-80 percent of them are alive at five years."
16
A study at Mayo Clinic evaluated the benefit of spiral CT and sputum cytology for early
detection of lung cancer in 1 520 patients. Fifty one lung cancers were detected in this study, and
59 percent were stage 1A. In addition to lung cancer, abdominal aneurysms, signs of
osteoporosis, and visceral fat were also detected. These findings suggest that full-body CT
screening will save lives through early detection of lung cancer and other diseases. 16
The Health Physics News has stated that "the amount of dose encountered from CT screening
with proper equipment is insufficient to present a legitimate concern of any radiation-related
health effects." 17
Radiation effects are frequently computed using the linear no-threshold model. This has been
called into question in many reports. 17 There may be no risks involved at low exposures, and
perhaps even benefits from radiation exposure. 18 Additional studies need to be conducted to
determine the benefits of whole-body CT screening. If legislation is passed prohibiting screening,
then the benefits in saved lives and reduced costs of care by early detection may never be known.
It is unreasonable to deny a benefit to someone because there may be an associated risk. Only if
the hazard is demonstrably greater than any benefit is it morally correct to deny a practice. The
risk is voluntary. We all assume acceptable risks in every activity we undertake.
There has been much thinking and evaluation about mammography and the appropriate signs for
a biopsy. The same should be done for CT screening follow up. Improvements in mammography,
in the face of a rising incidence of breast cancer, have contributed to a 20 percent reduction in
mortality.19 Total body and heart scanning can save lives by detecting diseases at earlier stages
when they are treatable. We do breast screening on women without symptoms. We should be
217
able to do the same with CT screening. There are recommendations as to when breast screening
should start and how often it should be done. We should develop the same for CT screening.
Rebuttal
Dr. Morin says the radiation risk of self-referral CT screening is slight. The issue is strictly a
risk-benefit analysis of how the findings are used. If the screening examination is not part of
preventative health care, then the findings can be considered as incidental to a normal healthcare
program. If these incidental findings cause hardship or excess cost, how does this differ from
other incidental findings? Are we to tell radiologists not to look for and report incidental findings
that occur during studies within the mainstream of current practice? If there is a problem with
self-referral findings, then there is also a problem with incidental findings in general.
As Dr. Morin points out, some states restrict the use of radiation screening. As he also points out,
however, the radiation risk is not really the issue here. Using radiation safety rules or arguments
to eliminate self-referral CT screening is disingenuous, and compromises the freedom of
individuals to make their own choices. There certainly will be some people who will benefit from
self-referral screening. There undoubtedly will be others who will not have acquired an overall
benefit. Professionals who want to play god and decide what is best for the public would be well-
advised to improve their profession's handling of the challenge of incidental findings, and leave
the risk-benefit judgment to an informed public. Although we would like to see truth in
advertising, the public is not as naive as the would-be protectors of the public's interests would
have us believe.
REFERENCES
1. "Survey of x-ray CT scanners," Conference of Radiation Control Program Directors,

www.crcpd.org
2. Florida Administrative Code, 64E-5.502.6
3. P.D. Bach et al., "Screening for lung cancer—A review of the current literature," Chest
123, 72S–82S (2003).
4. B.S. Kramer, J.K. Gohagan, and P.C. Prorok, Cancer Screening (Marcel Dekker, New
York, 1999).
5. G.S. Gazelle, "Screening for disease with whole body CT," Diagnostic Imaging
Conference on New Modalities and Radiation Safety, Honolulu, Hawaii, 2003.
6. L.F. Rogers, "Whole-body CT screening: Edging toward commerce," AJR, Am. J.
Roentgenol. 179, 823 (2002).
7. www.AAPM.org
8. www.ACR.org
9. www.CRCPD.org
10. www.fda.gov///cdrh/ct
11. S.J. Swensen et al., "Screening for lung cancer with low-dose spiral computed
tomography," Am. J. Respir. Crit. Care Med. 165, 508–513 (2002).
12. S.J. Swensen, "CT screening for lung cancer," AJR, Am. J. Roentgenol. 179, 833–836
(2002).
13. N.A. Obuchowski et al., "Ten criteria for effective screening: Their application to
multislice CT screening for pulmonary and colorectal cancers," AJR, Am. J. Roentgenol. 176,
1357–1362 (2001).
14. H.G. Welch, "Informed choice in cancer screening," JAMA, J. Am. Med. Assoc. 285,
2776–2778 (2001).
218
15. A. Jemal, T. Murray, A. Samuels, A. Ghafoor, E. Ward, and M.J. Thun, "Cancer
statistics, 2003," CA Cancer J. Clin. 53, 5–26 (2003).
16. RSNA News—November 2002.
17. AAPM President's Symposium, July 15, 2002.
18. M. Pollycove, "The rise and fall of the linear no-threshold (LNT) theory of radiation
carcinogenesis," Proceedings of Annual Meeting of the American Physical Society, Washington,
DC, 1997 April 21.
19. "ACR rebukes New York Times' portrayal of mammography," ACR Bulletin 58, 12–13
(2002).
219
5.5. It is time to retire the computed tomography dose index (CTDI)

for CT quality assurance and dose optimization
David J. Brenner and Cynthia H. McCollough

OVERVIEW
The computed tomography dose index (CTDI) was introduced over a quarter century ago for
optimization of patient protection in CT. 1 By means of a single measurement it was possible to
determine, to a good approximation, the average dose for a series of scans in lieu of making
multiple measurements for each slice. This advance made great sense at the time because of the
slow equipment and small anode heat capacities of early CT units, which made multiple
measurements difficult. It has recently been suggested that modern technological developments
in CT and dosimetry permit patient doses to be determined in a way that better represents the risk
to the patient, and that it is now time to retire the use of CTDI for CT quality assurance and dose
optimization.2 However, others argue that measurements of CTDI (or variants thereof) remain
adequate for CT quality assurance and dose optimization, and that replacement is unnecessary.
This difference of opinion is the topic of this month's Point/Counterpoint debate.
Arguing for the Proposition is David J. Brenner, Ph.D. Dr. Brenner is Professor of Radiation
Oncology and Public Health at the Columbia University Medical Center. He focuses on
developing models for the carcinogenic effects of ionizing radiation on living systems at the
chromosomal, cellular, tissue, and organism levels. He divides his research time roughly equally
between the effects of high doses of ionizing radiation (related to radiation therapy), and the
effects of low doses of radiation (related to radiological, environmental, and occupational
exposures). When not involved in radiation matters, he supports Liverpool Football Club.
Arguing against the Proposition is Cynthia H. McCollough, Ph.D. Dr. McCollough is Associate
Professor of Radiological Physics at the Mayo Clinic College of Medicine. She oversees the
technical support for Mayo's 22 CT scanners and directs the CT Clinical Innovation Center. Her
research interests include CT dosimetry, advanced CT technology, and new clinical applications.
She is an NIH-funded investigator, and is active in numerous organizations. She chairs the
AAPM's Task Group on CT Dosimetry and the ACR's CT Accreditation Physics Subcommittee,
and is a member of IEC, ICRU, and NCRP CT committees. Dr. McCollough received her
doctorate from the University of Wisconsin in 1991.
FOR THE PROPOSITION: David Brenner, Ph.D.
Opening Statement
We have an obligation to reduce, as far as practical, radiation-induced cancer risks in the

population who receive computed tomography (CT) examinations. These cancer risks are
determined by the organ doses to which individuals are exposed. It is logical for the quantities
measured in CT quality assurance and dose optimization (CT QA/DO) to bear as direct a
220
relationship to organ doses as is reasonably practical. The dose descriptors currently used for CT
QA/DO, the computed tomography dose index (CTDI) and its subsequent modifications,3 bear an
increasingly distant relationship to organ doses and thus to risk.4 The technology now exists to
directly, routinely and rapidly measure organ doses from helical CT scans in realistic
anthropomorphic phantoms, with about the same amount of technical effort as that required to
measure CTDI. Thus, I believe that such measurements represent a more logical basis for CT
QA/DO than do CTDI measurements.
Specifically, given 1) the problems in maintaining CTDI as a relevant dose index, 4 2) the
availability of MOSFET (Refs. 5,6) (or TLD, if preferred) dosimeters which are very small,
sensitive, quick, and convenient to use, and 3) the commercial availability of heterogeneous
whole-body anthropomorphic phantoms such as the ATOM phantoms7 and the Alderson
radiation therapy phantoms,8 it is time to consider retiring the CTDI/homogeneous phantom
approach to CT QA/DO. One might envisage CTDI measurements being replaced by direct
simultaneous MOSFET measurements of doses at locations in appropriate organs of a full-body
anthropomorphic phantom, perhaps appropriate subsets of stomach, colon, breast, lung, gonads,
thyroid, bladder, esophagus, liver, brain, and relevant bone marrow. A typical set of
measurements at 20 (simultaneously measured) organ locations should take about 30 minutes,
including setup—quite comparable to CTDI measurements.
There is no question that CTDI, and its related quantities, can be used to compare outputs of
different CT scanners and different CT models. But given the goal of minimizing unnecessary
cancer risks to patients, there is a need for a quantity that is a surrogate of risk, and neither CTDI
nor its modifications can be forced into this role. It is now quite practical to measure direct
surrogates for cancer risk, with no more technical effort than required to measure CTDI. It makes
sense to use these more direct measurements as the basis for CT quality assurance and dose
optimization.
AGAINST THE PROPOSITION: Cynthia McCollough, Ph.D.
Opening Statement
The advent of spiral CT caused concern about the use of a discrete axial scan to measure dose for
a continuous spiral acquisition. However, both theory and experimental data upheld the validity
of extending the CTDI construct to spiral CT.9,10 The larger problem, both for spiral or sequential
acquisitions, was the integration limits established in the early days of CT:±7 T, where T was the
nominal tomographic section thickness (in lay person language, the slice width). In the case of
narrow slice widths (which were not considered in the "early days"), the average dose from a
series of scans was underestimated by the ±7 T limits.11 Hence a fixed integration length of
100 mm, which purposely matched the active length of the well-established CTDI "pencil"
chamber, was adopted in Europe12 and in International CT Safety standards. 13 This resulted in a
CT dose index that is easily and reproducibly measured,3 and that captures the majority of the
scatter tails for even wide x-ray beam widths.14
Recently, the pitch-normalized metric CTDIvol was required by international standards to be

displayed on the user interface prior to scan initiation. 13 The radiology community, through
extensive educational efforts, is becoming "calibrated" to typical CTDI vol values for common
examinations, thereby allowing users to note scan prescriptions that deliver radiation levels
outside of the typical range. Users can use CTDIvol to provide a universal index of scanner output
221
that can be readily compared across scanners worldwide. This "apples to apples" comparison of
radiation doses in CT, where users can check scanner output prior to irradiation (and hopefully
modify techniques that are inappropriately different from the above reference values), is a
practical and robust method of dose optimization, as the use of reference values has consistently
been shown to reduce average dose levels and narrow the dose distribution across imaging
practices.15 CTDIvol is a valuable and necessary tool for this task, primarily because it is so well
established and uniformly adopted.
This uniformity in measurement technique makes CTDI an ideal quality assurance tool, as
quality assurance requires use of the same methods and phantoms in a consistent manner. So too,
does dose optimization. Knowing the dose to my liver or your liver is not the issue in clinical
dose optimization. Rather, one must know that a CTDIvol of 18–22 mGy is typical for an average
adult abdominal CT. That way, if a wrong parameter is selected leading to a CTDIvol of 59 mGy,
the user has a clear indication that something is wrong. Besides avoiding unnecessarily high dose
CT exams, the display of a universal, easily- and reproducibly-measured metric on the user
console provides the operator with a practical tool to reduce dose from CT examinations to
appropriate levels. Thus, I consider it time not to retire the CTDI, but rather to promote its use in
daily CT practice.
Rebuttal: David Brenner, Ph.D.
Professor McCollough cogently makes the point that if the sole object of the exercise is to
compare and confirm outputs from CT machines, as they are used in 2006, then the CTDIvol dose
index is just fine. There are several reasons, however, why CTDI is not the optimal way forward
for CT QA/DO.
First, if the history of CT dosimetry tells us anything, it is that the latest version of CTDI will
soon need to be modified due to changes in CT technology.3 For example, as multi-slice scanners
feature increasingly broad beams, the 100 mm ion chamber will no longer characterize enough
of the scatter from a single-slice profile.4,16 To have to base QA/DO on a dose index that needs to
be modified as CT technology changes is undesirable. Indeed, there are some imminent changes
in CT technology that are so basic that they cannot be accommodated by simply tinkering further
with CTDI. As an example, for continuous automated axial tube current modulation, designed to
compensate for changes in attenuation by different organs along the patient axis, 17 CTDI
measurements simply will not delineate whether or not the dose is being delivered appropriately
over the length of, say, a colon scan.
Secondly, Professor McCollough's central implication is that, in order to check that the scanner is
operating correctly, all we need for CT QA/DO is some good index of machine output. But if this
were so, even the basic CTDI100 would be more complicated than needed. In fact, still more
complicated, spatially-averaged versions of the CTDI100, like CTDIw and CTDIvol, are now the
standard.3 Why? Because they are slightly better surrogate indices for organ dose and thus
ultimately for organ risk!
In summary, there is a rationale and a desire in CT QA/DO to measure some quantity that will
need to be changed, and that is a better surrogate for organ dose/risk than is CTDI. So why not
directly measure organ doses in an anthropomorphic phantom?
Thus my arguments in support of the Proposition are:

222
1. Multiple organ dose measurements in an anthropomorphic phantom with a set of MOSFET

detectors, for example, are no harder or slower to make than CTDI measurements. 5,6
2. Organ dose measurements provide just as good a check that the machine is working correctly
as does CTDIvol.
3. The CTDI concept needs to be continuously modified as CT technology changes.
4. Organ dose measurements provide direct, rather than crude, surrogates of organ risk—the
quantity we ultimately want to control.
Rebuttal: Cynthia McCollough, Ph.D.
In CT, organ doses are determined by the start and end locations of the examination, scanner
output and patient anatomy. From the anatomy of interest, CTDI and scan protocol, organ doses
can be predicted with high precision using published Monte Carlo coefficients18,19 or Monte
Carlo code modified for this task. 20,21 Using "virtual phantoms" from actual patient CT scans,
dose optimization can easily be performed for patients of varying age, gender, and habitus for
countless perturbations of scan parameters. 22 The time, effort, and cost associated with "brute
force" measurements of organ doses for the innumerable combinations of detector
configurations, pitch values, kVp and mAs settings, beam shaping filters, and multiple child and
adult physical phantoms—per scanner model—is simply not practical. Further, physical
anthropomorphic phantoms, which are available in limited sizes, may use less-accurate
"geometric" organs, and can vary based on manufacturer or date of purchase. In addition to
dosimeter precision and calibration issues, such variations will confound the optimization task,
especially between investigators. Silicon-based dosimeters (diodes or MOSFETs) can only be
used on phantom surfaces (if placed internally, the wires create problematic gaps). Also, they
have spectral dependencies that are not easily addressed in CT, where spectra vary between
scanners and across the scan field, and they must be used in high-sensitivity mode for adequate
precision, which shortens their lifespan and increases user cost. TLDs, which can be placed
inside the phantom, require annealing and removal (to read them) between multiple
measurements—a time consuming effort. In contrast, CTDI gives a precise and consistent index
of scanner output, can be used to quickly measure output for many combinations of scanner
settings, and can be used with Monte Carlo tools for dose optimization. I agree that organ doses
are important, but physicists should use their time and talents to work smarter, not harder,
towards minimizing radiation risk from CT.
REFERENCES
1. R. Jucius and G. Kambic, "Radiation dosimetry in computed tomography CT," Proc.

SPIE 127, 286–295 (1977).
2. D. J. Brenner, "Is it time to retire the CTDI for CT quality assurance and dose
optimization?," Med. Phys. 32, 3225–3226 (2005).
3. M. F. McNitt-Gray, "AAPM/RSNA Physics Tutorial for Residents: Topics in CT.
Radiation dose in CT," Radiographics 22, 1541–1553 (2002).
4. R. L. Dixon, "A new look at CT dose measurement: beyond CTDI," Med. Phys. 30,
1272–1280 (2003).
5. M. W. Bower and D. E. Hintenlang, "The characterization of a commercial MOSFET
dosimeter system for use in diagnostic x ray," Health Phys. 75, 197–204 (1998).
223
6. D. J. Peet and M. D. Pryor, "Evaluation of a MOSFET radiation sensor for the

measurement of entrance surface dose in diagnostic radiology," Br. J. Radiol. 72, 562–568
(1999).
7. V. Varchenya, D. Gubatova, V. Sidorin, and S. Kalnitsky, "Children's heterogeneous
phantoms and their application in röntgenology," Radiat. Prot. Dosim. 49, 77–78 (1993).
8. S. W. Alderson, L. H. Lanzl, M. Rollins, and J. Spira, "An instrumented phantom system
for analog computation of treatment plans," Am. J. Roentgenol., Radium Ther. Nucl. Med. 87,
185–195 (1962).
9. M. F. McNitt-Gray et al., "Radiation dose in Spiral CT: The relative effects of
collimation and pitch," Med. Phys. 26(3), 409–414 (1999).
10. W. A. Kalender and A. Polacin, "Physical performance characteristics of spiral CT
scanning," Med. Phys. 18(5), 910–915 (1991).
11. American Association of Physicists in Medicine, Report No. 31, "Standardized Methods
for Measuring Diagnostic X-ray Exposures," 1990.
12. European guidelines for quality criteria for computed tomography (European
Commission, Luxembourg, 2000).
13. International Electrotechnical Commission, Medical Electrical Equipment. Part 2-44:
Particular requirements for the safety of x-ray equipment for computed tomography. IEC
publication No. 60601-2-44. Ed. 2.1. IEC Central Office: Geneva, Switzerland, 2002.
14. C. H. McCollough, F. E. Zink, and R. L. Morin, "Radiation dosimetry for electron beam
CT," Radiology 192, 637–643 (1994).
15. D. Hart et al., "Doses to patients from medical x-ray examinations in the UK–1995
review," Health Protection Agency, Chilton, UK: NRPB-R289 (1996).
16. S. Mori, M. Endo, K. Nishizawa, T. Tsunoo, T. Aoyama, H. Fujiwara, and K. Murase,
"Enlarged longitudinal dose profiles in cone-beam CT and the need for modified dosimetry,"
Med. Phys. 32, 1061–1969 (2005).
17. M. K. Kalra, M. M. Maher, T. L. Toth, B. Schmidt, B. L. Westerman, H. T. Morgan, and
S. Saini, "Techniques and applications of automatic tube current modulation for CT," Radiology
233, 649–657 (2004).
18. D. G. Jones and P. C. Shrimpton, Report No. NRPB-SR250, Health Protection Agency,
Chilton, UK (1993).
19. M. Zankl, W. Panzer, and G. Drexler, "The calculation of dose from external photon
exposures using reference human phantoms and Monte Carlo methods Part VI: Organ doses
from computed tomographic examinations," GSF-Bericht 30/91. Neuherberg, Germany:
Gesellschaft für Strahlen- und Umweltforschung (1991).
20. W. A. Kalender et al., "A PC program for estimating organ dose and effective dose
values in computed tomography," Eur. Radiol. 9 (3), 555 (1999).
21. J. J. DeMarco et al., "A Monte Carlo based method to estimate radiation dose from
multidetector CT (MDCT).: cylindrical and anthropomorphic phantoms," Phys. Med. Biol. 50
(17), 3989 (2005).
22. N. Petoussi-Henss et al., "The GSF family of voxel phantoms," Phys. Med. Biol. 47 (1),
89 (2002).
224
5.6. A major advantage of digital imaging for general radiography is

the potential for reduced patient dose so screen/film systems should
be phased out as unnecessarily hazardous
Eliseo Vano and Keith Faulkner

OVERVIEW
In general radiography, the dose to the patient is so low that any biological risks to the exposed
individual are considered negligible compared with the benefit received. Because of the
enormous number of procedures performed, however, the collective dose to the entire population
of exposed individuals is significant. Many organizations, such as the International Commission
on Radiological Protection (ICRP), recommend keeping this collective dose as low as possible. 1
Since digital detectors have higher intrinsic radiation sensitivities compared with film/screen
systems, one way to achieve this objective may be to replace film/screen radiography with digital
imaging. Whether this concept is viable in practice is the premise for this month's
Point/Counterpoint debate.
Arguing for the Proposition is Eliseo Vano, Ph.D. Dr. Vano is Professor of Medical Physics at
the Medical School of the Complutense University, and Head of the Medical Physics Service at
the San Carlos University Hospital in Madrid, Spain. After graduation in physics at the
University of Valencia, he worked in Nuclear Spectroscopy in Madrid and Paris, and obtained
his Ph.D. degree at the Complutense University of Madrid (CUM). Dr. Vano created a Medical
Physics Group at the Medical School in the CUM and, in 1982, became the first Professor of
Medical Physics in Spain. He contributed to introduction of quality control and patient protection
procedures in diagnostic radiology in Spain, is the current Secretary of Committee 3 of the ICRP
(Protection in Medicine), and Vice Chairman of the Medical Working Party of the Group of
Experts of the Euratom Treaty. He has published extensively in diagnostic and interventional
radiology and cardiology, on topics such as quality control and radiation protection of patients
and workers.
Arguing against the Proposition is Keith Faulkner Ph.D. Dr. Faulkner is the Regional Director of
Quality Assurance for the breast and cervical cancer screening programs for the North East,
Yorkshire and Humber Government Office Regions of the UK. After graduation from Imperial
College, London, he trained as a medical physicist at the University of London. He is a Fellow of
the Institute of Physics and Engineering in Medicine (IPEM), the Institute of Physics, and the
Society of Radiation Protection (SRP) in the United Kingdom. He maintains an active interest in
research in radiation protection, and is the coordinator of the European Commission's
SENTINEL project. He is the author of over 250 peer-reviewed papers. His awards include the
Founders' Prizes of the IPEM and the SRP, and the Barclay Prize and Medal of the British
Institute of Radiology.
FOR THE PROPOSITION: Eliseo Vano, Ph.D.

225
Opening Statement
With appropriate training of users and relevant quality control, it is possible to generate images
with the same or lower radiation dose with digital imaging compared with conventional
film/screen systems for general radiography.1 In addition, image quality and diagnostic
information (after post-processing) can be improved with digital techniques. 2,3,4 Transmission of
images by networks and easy archiving are other fundamental advantages of digital imaging.
Networks and digital systems also enable the auditing of patient doses and comparison with
reference values in real time.5 With film/screen systems, errors in radiographic techniques or
problems with film processors require repeat exposures whereas, with digital imaging, the
inherent wide dynamic range and post-processing capability make most retakes unnecessary.
Furthermore, if previous films are not available, additional exposures may be required. Hence the
use of film/screen systems should be phased out within the next few years because they are
unnecessarily hazardous and less versatile than digital imaging.
Digital radiology can have additional benefits for developing countries. Digital imaging can be
less expensive than film/screen techniques. Compared with film processors, the use and
maintenance of digital phosphor plates (computed radiology) and personal computers (PCs) is
easier. Even if initial funding to install sophisticated networks and image archiving systems is not
available in developing countries, users can use PCs to burn CD-ROMs that are cheaper than
radiographic films and more easily transported by patients to other hospitals, if necessary.
With film/screen systems, patient doses at the individual level are not so high as to be considered
hazardous. However, if we take into account the collective dose for a medium sized hospital with,
say, 100,000 examinations per year, and if 30% of the patient dose could be saved with digital
techniques and another 10% by avoidance of retakes, a substantial risk could be avoided and
potential cost savings could be realized. With a typical mean effective dose per examination of
0.5 mSv (collectively, 50,000 mSv/year), digital imaging, if used properly, could reduce the
annual collective dose by 20,000 mSv. This annual reduction in collective dose for the same
clinical benefit per hospital, supports the promotion and introduction of digital technology as fast
as possible.
The initial effort to train the staff in the use of digital techniques and the development of the
necessary quality control program would be offset by the dose reduction and the workflow
improvement.
AGAINST THE PROPOSITION: Keith Faulkner, Ph.D.
Opening Statement
There are many valid reasons for introducing digital imaging into general radiography, but the
claim that film/screen systems are unnecessarily hazardous is not one of them. Digital imaging
for general radiography has many advantages compared with film/screen systems, including a
wider dynamic range and the ability to manipulate the images produced. The wider dynamic
range means that acceptable images may be acquired at a range of dose levels, and therefore
repeat exposures can be reduced. Digital imaging can result in the over use of radiation, however,
because there is a tendency for images to be acquired at too high a dose, and for too many images
to be obtained, simply because the acquisition process is so easy.
226
There are two aspects of the Point/Counterpoint proposition that can be challenged. The first is
that digital imaging necessarily reduces patient doses, and the second is that film screens should
be phased out as unnecessarily hazardous. Each of these is discussed below.
In digital imaging, increasing patient dose generally means improving image quality because of
reduced image noise. Digital imaging systems may be set up to produce higher quality images
than necessary for diagnosis,6 and consequently patient doses may be higher than needed. The
DIMOND project has proposed that digital radiography procedures should be classified in low,
medium or high image quality bands.7 Advice is given on the speed class of the imaging
modality to be used for each of these three bands. In these proposals, the speed class of a
film/screen system for a given image quality band is the same as that of a storage-phosphor
system, and one band lower than that for flat-panel detectors. The decrease in patient dose
associated with the introduction of a flat-panel system for digital radiology lies in the range of
33–50% when replacing a film/screen system.8,9,10 Thus, even if flat-panels are used properly, the
maximum reduction in patient dose is only a factor of two. Hence, screen/film systems should not
be considered unnecessarily hazardous.
The ease of image acquisition by digital systems can mean that there is an increase in the number
of images acquired. This has happened in digital fluoroscopy.11 Digital radiology has also led to
an unjustified increase in the number of procedures by 82%.12
In summary, there is evidence that the introduction of flat-panel detectors leads to a dose-
reduction of typically 50%, but can lead to almost a doubling of the number of procedures.
Hence, it is not reasonable to conclude that film/screen systems are unnecessarily hazardous, and
the Point/Counterpoint proposition is untenable. The International Commission on Radiological
Protection also has concerns about dose levels in digital radiology. These concerns are
summarized in Publication 93.1
Rebuttal: Eliseo Vano, Ph.D.
I agree in part with some of the points presented by Dr. Faulkner.
The wide dynamic range of digital detectors means that acceptable images may be acquired at a
range of dose levels and that overexposures may not be detected. But overexposures will be
minimized with proper training programs and quality control, including frequent patient dose
evaluations.
In digital imaging increased patient dose generally means improved image quality due to reduced
noise. Again, I agree that this could be misused if the required image quality is not defined for a
specific medical task. With digital imaging, it is possible to select image quality (and radiation
dose) using the same detector, which is not the case with film/screen. With this already "old"
imaging technology, it is necessary to change the speed class of the film/screen combination and,
in practice, more dose than is necessary is sometimes used.
The example quoted by Dr. Faulkner of the European DIMOND project 7 proposing the model of
three image-quality bands in projection radiography (and corresponding dose levels of low,
medium and high) is a clear advance in patient dose management. The ability to select low image
quality for some clinical tasks, leading to a saving of 50% in patient dose, is not so readily
available if film/screen systems are used.
227
Dr. Faulkner points out the tendency in digital radiology to obtain more images than necessary
(especially in fluoroscopy), but this again is a matter of appropriate training of the practitioners.
Using a digital system and archiving short fluoroscopic sequences, a well-trained radiologist may
complete a digestive or urological examination, for example, with fewer images and less patient
dose compared with those for non-digital technology.
I agree with Dr. Faulkner that film/screen systems are not "per se" unnecessarily hazardous, but it
is clear that the added benefits of digital technology are enormous. These advantages, together
with reduced doses, means that we should advise our health administrators to move in this
direction.
Rebuttal: Keith Faulkner, Ph.D.
Professor Vano makes a series of important points in supporting the use of digital imaging for
general radiography. I wholeheartedly agree with him that digital imaging should result in lower
patient doses, provided that the technology is used appropriately. I concur that one of the main
benefits of digital imaging is the large dynamic range compared with that for film/screen
systems. The wide dynamic range means that it is possible to obtain acceptable images at a range
of dose levels, reducing the need for repeat exposures. However, there is a tendency for digital
systems to be used at higher dose levels than necessary, as users demand images of the highest
quality rather than of a quality needed for diagnosis. Professor Vano is correct in endorsing the
role of training; it is critical that users have an understanding of the technology of the system, if
the potential for dose reduction associated with the use of digital technology is to be achieved in
practice.
Even if it is accepted that digital imaging can reduce doses by up to a factor of two, this level of
dose reduction does not imply that film/screen systems are unnecessarily hazardous. There are
many areas of radiology practice where local differences result in dose variations that are much
greater than two, and several population surveys have shown that doses vary among centers by
factors of 10–100.
Placed in this context, film/screens should not be regarded as unnecessarily hazardous. There are
valid reasons for replacing film/screen systems with digital detectors for general radiography,
particularly for centers where there may be problems with the water supply that could impact on
chemical processing of film. Professor Vano has eloquently made the case for change, but the
case should not be based upon the proposition that film/screens are unnecessarily hazardous.
REFERENCES
1. International Commission on Radiological Protection. ICRP-93. "Managing patient dose

in digital radiology. A report of the International Commission on Radiological Protection," Ann.
ICRP 34, 1–73 (2004).
2. U. Neitzel, "Status and prospects of digital detector technology for CR and DR," Radiat.
Prot. Dosim. 114, 32–38 (2005).
3. J. Persliden, E. Helmrot, P. Hjort, and M. Resjo, "Dose and image quality in the
comparison of analogue and digital techniques in paediatric urology examinations," Eur. Radiol.
14, 638–644 (2004).
4. P. Monnin et al., "Performance comparison of an active matrix flat panel imager,
computed radiography system, and a screen-film system at four standard radiation qualities,"
Med. Phys. 32, 343–350 (2005).
228
5. E. Vano et al., "Real-time measurement and audit of radiation dose to patients

undergoing computed radiography," Radiology 225, 283–288 (2002).
6. S. Busch, C. Decker, C. Schilz, and H. P. Busch, "Image quality and exposure dose in
digital projection radiography," Fortschr. Röntgenstr. 175, 32–37 (2003).
7. H. P. Busch, "Management of dose and quality in digital radiography," Fortschr.
Röntgenstr. 175, 17–19 (2003).
8. M. Strotzer et al., "Routine chest radiography using a flat-panel detector: image quality at
standard detector dose and 33% dose reduction," Am. J. Roentgenol. 178, 169–171 (2002).
9. C. Fink et al., "Clinical comparative study with a large area amorphous silicon flat-panel
detector: image quality and visibility of anatomic structures in chest radiography," Am. J.
Roentgenol. 178, 481–486 (2002).
10. K. Herrmann et al., "Chest imaging with flat-panel detector at low and standard doses:
comparison with storage phosphor technology in normal patients," Eur. Radiol. 12, 385–390
(2002).
11. B. Axelsson et al., "A comparison of analogue and digital techniques in upper
gastrointestinal examinations: absorbed dose and diagnostic quality of the images," Eur. Radiol.
10, 1351–1354 (2000).
12. B. Reiner et al., "Effects of filmless imaging on the utilization of radiologic services,"
Radiology 215, 163–167 (2000).
229
5.7. Within the next five years CT colonography will make

conventional colonoscopy obsolete for colon cancer screening
Hiro Yoshida and Albert C. Svoboda, Jr.

OVERVIEW
In the United States colorectal cancer is second only to lung cancer as a cause of cancer death,
despite the fact that screening by colonoscopy can detect colorectal cancer early enough for cure
and, by detection and removal of precancerous polyps, can actually prevent cancer.1 The major
reason for this high cancer death rate is that fewer than 20% of at-risk adults are screened.1
Unfortunately there is not sufficient capacity to screen the entire population at risk, and it has
been estimated that it would take 10 years to provide colonoscopies to all these unscreened
individuals.2 Sufficient capacity does exist, however, for screening with CT colonography, which
has been demonstrated to be a good alternative to colonoscopy but does not suffer from the "fear
factor" associated with having an invasive procedure under sedation. 3 Because of these
advantages, it seems possible that CT colonography might soon replace colonoscopy as the
screening method of choice, and this is the premise debated in this month's Point/Counterpoint.
Arguing for the Proposition is Hiro Yoshida, Ph.D. Dr. Yoshida received his B.S. and M.S.
degrees in physics and a Ph.D. degree in information science from the University of Tokyo. He
previously held an Assistant Professorship in the Department of Radiology at the University of
Chicago. He was a tenured Associate Professor when he left the university and joined the
Massachusetts General Hospital and Harvard Medical School in 2005, where he is currently an
Associate Professor of Radiology. His research interest is in computer-aided diagnosis, in
particular the detection of polyps in CT colonography, for which he received several awards from
the Annual Meetings of the RSNA.
Arguing against the Proposition is Albert C. Svoboda, Jr., MD. Dr. Svoboda received a B.S.
(Zoology) (1955) and an MD (1958) from the University of Chicago as well as an M.S.
(Physiology) (1954) from the University of Southern California, and completed his internship,
residency and fellowship in gastroenterology at the University of Michigan (1958–1963).
Subsequently, he became an Associate in Gastroenterology at the Scripps Clinic and Research
Foundation (1963–1966) and the Department of Gastroenterology, Sansum Medical Clinic
(1966–1996). He is currently a Senior Scientist at the Sansum Diabetic Research Institute and
Director of the Internship Program. In the American College of Gastroenterology he has served
as the Governor for Northern California, Secretary, and Chair of the Board of Governors. He is a
Past President of the Southern California Society for GI Endoscopy. In his spare time, Dr.
Svoboda is an ardent orchid enthusiast and is an Accredited Judge for the American Orchid
Society and the Cymbidium Society of America.
FOR THE PROPOSITION: Hiro Yoshida, Ph.D.
Opening Statement
230
Colonoscopy is considered the most accurate test for the detection of colonic neoplasia, and it has
recently become a popular screening tool for high-risk patients.4 However, current screening
techniques available to average-risk patients are many, and those who make screening
recommendations do not agree as to which test or combination of tests should be performed.5,6
Thanks to the rapid advancement of computer tomographic (CT) technology, CT colonography,

popularly known as virtual colonoscopy, has been emerging as a promising alternative technique
for screening.1 CT colonography involves performing an abdominal CT scan to create images
resembling those seen at colonoscopy for detecting colorectal neoplasms. As CT colonography
evolves, the weaknesses of colonoscopy will be exposed.
Colonoscopy is an invasive intervention that carries the risk of perforation or serious

hemorrhage,7 whereas CT colonography is a noninvasive image-based examination of the entire
colon, with no report of complications thus far. Furthermore, screening of the entire at-risk
population by colonoscopy may not be realistic: approximately 41.8 million average-risk
Americans 50 years old or older have not been screened, and 12 million more reach 50 each year.
Current and near-future capacity for colonoscopy is severely limited and will be unable to satisfy
this huge demand.2 Cost-effectiveness comparison of colonoscopy and CT colonography is
inconclusive; however, it indicates that if the cost of virtual colonoscopy could be reduced to
approximately half that of colonoscopy, which is not unrealistic, CT colonography would become
the more cost-effective screening tool.8
Advocating CT colonography obviously requires formal evidence-based scientific data on its

ability to detect polyps. A recent large clinical trial, in which 1,233 asymptomatic average-risk
patients were examined by state-of-the-art CT colonography, showed a per-polyp sensitivity of
94% and a specificity of 92% for adenomatous polyps >8 mm.9 This sensitivity of CT
colonography was higher than that of colonoscopy (91%), and this is consistent with the reported
average miss rate of 6–13% in colonoscopy.10 Critics argue that the high performance in the
study can be achieved only in ideal circumstances. However, in the next five years, the
performance of CT colonography is expected to improve further by use of rapidly advancing new
technologies such as computer-aided detection11,12 and novel views.13
CT colonography is now showing results that may lead to its becoming a mainstream screening
test, which will make colonoscopy obsolete as the primary screening tool. However, it will not
eradicate colonoscopy. Colonoscopy has two separate functions, a diagnostic procedure and a
therapeutic technique for removing polyps. A realistic scenario is to perform colonoscopy only
selectively, for patients with polyps of clinically significant size detected by primary CT
colonography. This scenario would be a more efficient use of endoscopy resources, and it is
likely to reduce the incidence of colorectal cancer.8
AGAINST THE PROPOSITION: Albert C. Svoboda, Jr., MD
Opening Statement
"Be not the first by whom the new are tried,
Nor yet the last to lay the old aside."
An Essay on Criticism, Part II, Alexander Pope (1688–1744)

231
"A bird in the hand is worth two in the bush;
A pound in the purse is worth two in the book."
My seniority brings with it "Remembrances of Times Past" when colon polyps or cancers were
detected by single column or air contrast barium enema. Perplexing then as today with
colonography was the question of what to do with the smooth polyp of small or moderate size,
say 1 cm. Prevailing wisdom then was to repeat the x-ray examination after three months and, if
the polyp had increased in diameter by 50%, to offer the patient surgery. Today with
colonography are the majority of these patients sent to surgery (e.g., colonoscopy)? Would the
high risk patient (strong family or personal history) be better served by referral directly for
colonoscopy?
Colonoscopy is clearly superior for therapeutic intervention and pathologic diagnosis, but my
task is to evaluate and predict diagnostic improvements.
"And differing judgements serve but to declare
The truth lies some where, if we knew but where."
William Cowper (1731–1800)
In recent insights on endoscopic innovations, 14 Dr. P. Jay Pasricha (UTMB, Galveston, Texas)
said that any new procedure should (1) meet a large need, (2) be simple and safe, and (3) be of
proven efficacy. Colonography must answer questions of radiation safety and ability to detect flat
lesions, answer follow-up concerns about small polyps, and try to match the extensively proven
safety and efficacy of colonoscopy.
Since the introduction of fiberoptic colonoscopy by Bergein Overholt in 1963, followed by

cancer detection in 1966, and endoscopic polypectomy by Wolff and Shinya in 1969, there have
been steady improvements in diagnostic and therapeutic capabilities of the instruments, which
still continue. In addition to video endoscopy, chromoendoscopy and high magnification,
improvements can be expected to mirror the on-going progress of miniaturization and optics.15
During these decades the procedure has been proven to be both safe and effective even in the
hands of increasing numbers of practitioners with variable training. Both colonoscopy and
flexible sigmoidoscopy have been shown to significantly and statistically reduce both the number
of cancers and deaths from cancer in the populations followed. Recognizing this, Medicare and
most insurance companies cover the costs of colorectal cancer screening including appropriate
endoscopy.
Any new procedure advocated to replace colonoscopy must demonstrate equal or greater safety,
efficiency, availability, affordability, and patient acceptance.
"People rather admire what is new although they do not know whether it be proper or not, than
what they are accustomed to and know already to be proper; and what is strange, they prefer to
what is obvious."
Hippocrates
232
Rebuttal: Hiro Yoshida, Ph.D.
As Dr. Svoboda correctly pointed out, any new procedure intended to replace colonoscopy must
demonstrate equal or greater safety, efficiency, availability, affordability, and patient acceptance.
Current 16-slice multi-detector CT scanners allow efficient acquisition of CT images of the colon
in 15–20 seconds with a minimum of patient discomfort and invasiveness. The success of low-
dose CT colonography16,17 is removing the concern about radiation safety. The rapid increase in
installed CT scanners makes CT colonography widely available to patients. Advances in
visualization and computer-aided detection software are expected to permit efficient
interpretation while maintaining high detection accuracy, making CT colonography cost
effective. Once Medicare begins to reimburse for screening CT colonography, the procedure will
become affordable for elderly patients whose main risk factor is age and who constitute the
majority of average-risk patients. The emergence of "prepless" CT colonography with fecal
tagging18 is removing the need for laxatives—the foremost reason for the low patient compliance
in colonoscopy and CT colonography—thus increasing patients' acceptance. Unfortunately, the
detection of flat lesions is as difficult for CT colonography as it is for colonoscopy. 19
Rebuttal: Albert C. Svoboda, Jr., MD
To allow an honest debate we must not compare apples and oranges. We cannot compare
complications from polypectomy and thermo-elective therapy with screening colonography. We
cannot ignore additive radiation effects from repeated surveillance virtual colonoscopy nor
ignore the neoplastic potential of undetected or untreated flat or small polypoid lesions. And we
cannot extend the good but unmatched detection rate of a single unit to general usage.
The introduction of colonography has been a clarion call for colonoscopists to review and
compare rates of complete exams (to cecum), missed lesions and subsequent development of
neoplasms.20,21 Both flexible sigmoidoscopy and colonoscopy have a long history of careful
follow-up of patients (e.g., the National Polyp Study) and proven reduction of cancer risk.
A recent review of colonoscopies by a gastroenterology group in Rockford, Illinois21 should set

achievable performance standards and endoscopy goals. All six endoscopists were experienced
(had done more than 2000 cases). The cecum was reached in a minimum of 98% of cases by
each. Of 16,147 colonoscopies performed there were no perforations from diagnostic procedures.
Of 206 patients with colon cancer, eight (3.8%) had a colonoscopy between six and thirty-six
months previously and five of these occurred at the site of the previous polypectomy suggesting
incomplete removal. The remaining three, all of which were early stage, apparently were missed.
One disadvantage of CT colonography is the radiation exposure required. Although the risk to
the individual patient is very low, the cumulative risk to the entire population of screened patients
is of some concern.22 Also, many patients will avoid CT because of their fear of radiation, even
though such fear might be irrational.
Dr. Yoshida mentions the risk of complications. I performed many thousands of colonoscopies
over thirty years of clinical practice. There were no deaths and very few perforations (less than
0.05%), all of which were related to difficult therapeutic procedures. The complications for
colonoscopy quoted by Dr. Yoshida are virtually all related to therapeutic endoscopy and the use
of thermal devices. Screening colonoscopy should be essentially free of complications especially
in the hands of trained, experienced endoscopists.21 The technique of cold snaring of small
lesions similarly carries almost no risk, probably no different than the air insufflation for CT
233
colonography. Careful patient monitoring for oxygen saturation, heart rate, and blood pressure
should avoid problems from conscious sedation.
The most common cause of medical litigation is failure to diagnose. This is almost as critical in
the physician's thinking as the patient's well being. If the criterion for referral for colonoscopy is
the detection of polyps as small as 6 mm, four times as many procedures will be necessary than
if the lower limit is 1 cm. With a 6 mm limit the false positive rate observed is of the order of
55%. A high false positive rate reduces the benefit of pre-selection offered by virtual
colonoscopy in the first place. The number of patients who will require both procedures could be
a budget breaker. There is pressure to refer patients for colonoscopy only if larger polyps are
found. If you see small polyps and report these but do not refer the patient for colonoscopy and
cancer develops, you are liable.
In terms of cost effectiveness, Sonnenberg et al.,23 using a Markov model, concluded that CT
colonography would be more cost effective than conventional colonoscopy only if the cost of CT
could be reduced to 50% of the cost of colonoscopy. Can CT colonography compete?
Both optical and virtual colonoscopy are evolving rapidly and improving. I end with the final
thought: "It is not the strongest of the species that survives nor the most intelligent; it is the one
most adaptable to change."
Charles Darwin (1805–1882)
REFERENCES
1. T. Byers et al., "American Cancer Society guidelines for screening and surveillance for
early detection of colorectal polyps and cancer: update 1997," American Cancer Society
Detection and Treatment Advisory Group on Colorectal Cancer, Ca-Cancer J. Clin. 47, 154–160
(1997).
2. L. C. Seeff et al., "Is there endoscopic capacity to provide colorectal cancer screening to
the unscreened population in the United States?," Gastroenterology 127, 1661–1669 (2004).
3. J. T. Ferrucci, "Colon cancer screening with virtual colonoscopy: promise, polyps,
politics," Am. J. Gastroenterol. 177, 975–988 (2001).
4. D. K. Rex, "American College of Gastroenterology action plan for colorectal cancer
prevention," Am. J. Gastroenterol. 99, 574–577 (2004).
5. G. S. Gazelle, P. M. McMahon, and F. J. Scholz, "Screening for colorectal cancer,"
Radiology 215, 327–335 (2000).
6. D. F. Ransohoff, "Colon cancer screening in 2005: status and challenges,"
Gastroenterology 128, 1685–1695 (2005).
7. N. M. Gatto et al., "Risk of perforation after colonoscopy and sigmoidoscopy: a
population-based study," J. Natl. Cancer Inst. 95, 230–236 (2003).
8. U. Ladabaum, K. Song, and A. M. Fendrick, "Colorectal neoplasia screening with virtual
colonoscopy: when, at what cost, and with what national impact?" Clin. Gastroenterol. Hepatol.
2, 554–563 (2004).
9. P. J. Pickhardt et al., "Computed tomographic virtual colonoscopy to screen for colorectal
neoplasia in asymptomatic adults," N. Engl. J. Med. 349, 2191–2200 (2003).
10. D. K. Rex et al., "Colonoscopic miss rates of adenomas determined by back-to-back
colonoscopies," Gastroenterology 112, 24–28 (1997).
11. H. Yoshida and A. H. Dachman, "CAD techniques, challenges, and controversies in
computed tomographic colonography," Abdom. Imaging 30, 26–41 (2005).
234
12. J. Näppi and H. Yoshida, "Feature-guided analysis for reduction of false positives in
CAD of polyps for computed tomographic colonography," Med. Phys. 30, 1592–1601 (2003).
13. C. F. Beaulieu, D. S. Paik, S. Napel, and R. B. Jeffrey, "Advanced 3D Display Methods,"
in Atlas of Virtual Colonoscopy, edited by A. H. Dachman (Springer-Verlag, New York, 2003),
pp. 37–44.
14. P. Jay Pasricha, "Innovations in Endoscopy," American Gastroenterological Association
(http://www.agacmecafe.com/agacmecafe/endoscopy.asp).
15. D. K. Rex, V. Chadalwada, and D. S. Helper, "Wide angle colonoscopy with a prototype
instrument: impact on miss rates and efficiency as determined by back to back colonoscopies,"
Am. J. Gastroenterol. 98, 2000–2005 (2003).
16. M. Macari et al., "Colorectal neoplasms: prospective comparison of thin-section low-dose
multi-detector row CT colonography and conventional colonoscopy for detection," Radiology
224, 383–392 (2002).
17. R. Iannaccone et al., "Colorectal polyps: detection with low-dose multi-detector row
helical CT colonography versus two sequential colonoscopies," Radiology 237, 927–937 (2005).
18. R. Iannaccone et al., "Computed tomographic colonography without cathartic preparation
for the detection of colorectal polyps," Gastroenterology 127, 1300–1311 (2004).
19. S. H. Park et al., "False-negative results at multi-detector row CT colonography:
multivariate analysis of causes for missed lesions," Radiology 235, 495–502 (2005).
20. J. C. Van Rijn et al., "Polyp miss rates determined by tandem colonoscopy. A systematic
review," Am. J. Gastroenterol. 101, 343–350 (2006).
21. J. F. Johanson and S. Burke, "Quality assessment of colonoscopy—colonoscopy miss rate
for colon cancer in the private practice," Abstract 7, ACG 2005 Annual Scientific Meeting.
22. D. J. Brenner and M. A. Georgsson, "Mass screening with CT colonography: should the
radiation exposure be of concern?," Gastroenterology 129, 328–337 (2005).
23. A. Sonnenberg, F. Delco, and P. Bauerfeind, "Is virtual colonoscopy a cost-effective
option to screen for colorectal cancer?," Am. J. Gastroenterol. 94, 2269–2274 (1999).
235
5.8. For diagnostic imaging film will eventually be of historical

interest only
Brent K. Stewart and Frank N. Ranallo

Reproduced from Medical Physics, Vol. 26, No. 5, pp. 669-671, May 1999
OVERVIEW
Twenty years ago some radiology savants predicted that film would disappear from radiology
over the next decade and was to be replaced by digital technologies for acquiring, storing, and
displaying images. Today some (principally academic) radiology departments are committed to
this objective, and a few (primarily those in military and VA hospitals that have access to the
Federal largesse) have achieved it. But most departments are adopting a ‗‗wait-and-see‘‘ attitude,
while using ‗‗mini-PACS‘‘ systems to bridge between specific digital imaging systems. Often
individuals in these departments believe that the up-front cost of going ‗‗all digital‘‘ will prevent
its widespread adoption. They emphasize that if radiology had evolved digitally, and film were a
new discovery, it would be heralded as a major contribution to the quality and cost-effectiveness
of medical imaging.
How this controversy is likely to be resolved is the subject of this Point/Counterpoint issue.
Arguing for the Proposition is Brent K. Stewart. Dr. Stewart received a Bachelor of Science in
physics from the University of Washington and a Ph.D. in Biomedical Physics from the
University of California, Los Angeles, where he was a National Science Foundation fellow. He
has over 140 peer-reviewed journal articles, book chapters, proceedings papers, and abstracts. He
has been involved in Picture Archiving and Communication Systems (PACS) research for the
past fifteen years. He is currently Professor of Radiology, Bioengineering, and Medical
Education as well as Director of Imaging Informatics and Director of Diagnostic Physics at the
University of Washington School of Medicine in Seattle.
Arguing against the proposition is Frank N. Ranallo. Dr. Ranallo is a clinical medical physicist at
the University of Wisconsin Hospital and Clinics and an Assistant Professor (CHS) in the
Department of Medical Physics at the University of Wisconsin. He received his Ph.D. in Physics
from the University of Wisconsin, and has been working in the field of medical imaging since
1976. He is co-author of a book on MRI and has served in the AAPM as a member of the
Diagnostic X-Ray Imaging Committee and its task groups. Dr. Ranallo‘s research interests
include optimization of the design and use of imaging equipment and the development of
instrumentation and methods to test the performance of medical imaging systems.
For the proposition: Brent K. Stewart
Opening Statement
Since Roentgen‘s day, film has been the traditional method for capturing and displaying
information in medical imaging. This method has served the discipline well; filmscreen
radiography can achieve exquisite spatial resolution and provides a reasonable dynamic range of
236
optical densities for display of image contrast. Film is a jack of all trades: It serves as image
acquisition, display, and storage devices. No other medium can compare with film‘s multitasking
capability, even though film is by no means optimized for any one task.
The greatest attraction of film is that it is analog. We live in an analog world; our everyday
experience is continuous, not discrete. Film is very user friendly. We humans are very tactile
creatures, and touching makes objects more ‗‗real‘‘ to us. There is nothing in the digital world
quite like the solid feel of paging through a good book or passing family snapshots around the
kitchen table. Even though I digitize almost all messages crossing my desk, I still print many out
for meetings or if there are several complex documents that I need to refer to. There is no
paperless office yet, but we are inching towards it daily.
However, film has some severe disadvantages that are becoming increasingly problematic in a
busy radiology service. These disadvantages include: the frequency of repeat exposures; films
are not immediately available for review; it is necessary to properly dispose of the by-products of
film processing; and the inability to spontaneously view a film simultaneously anywhere within
an institution, the country or the world. The film library is often the bottleneck in the clinical
information workflow constituting an interpretive and consultative medical imaging service.
Lost, misplaced and misappropriated films are a constant problem that not only frustrates
radiologists and referring physicians, but also compromises the quality of patient care. We will
never have a ‗‗cradle to grave‘‘ medical record using paper and film.
The widespread adoption of electronic radiology (aka PACS, IMACS, etc.) is inevitable, just as
electric lighting, radio, and television have been. All new technologies, when invented,
developed and marketed go through an adoption life cycle. The percent of society adopting the
technology over time generally describes a sigmoidal-shaped curve, ironically much like an
H&D curve. The evolution of technology adoption follows a fairly predictable course. First the
Innovators (2%–4%) break new technological ground. Next come the Visionaries, or the Early
Adopters (12%–14%). Between the Early Adopters and the Early Majority lies the ‗‗chasm,‘‘
where many technological innovations become mired. The PACS chasm has been bridged lately
and the Early Majority (34%) has begun implementing electronic radiology. The Late Majority
(34%) and the Laggards (16%) bring up the rear. We are not too far along the technology
adoption curve for PACS, but the next few years will see rapid deployment of these systems.
Eventually electronic radiology will be fully adopted as the standard for patient care, the
Laggards will finally capitulate, and film will be relegated to the silver reclamation bin.
Rebuttal
While digital vanguards have been lauding the demise of radiographic film for the past twenty
years, the process of filmless radiology has taken longer than expected due to the economic,
technological, and human elements involved. In 1983, while a graduate student working on the
seminal elements of PACS at UCLA, filmless radiology seemed relatively straightforward. There
were many strong and clear motivations. We have, however, had to wait several years for cost-
effective technology to catch-up with the vision of PACS. There is much unfinished work to
bring filmless radiography to fruition, including information technology solutions that can adapt
their functionality based on user work-flow preferences.
The driver for filmless radiology is not anticipated cost or film library space savings, but the
economic imperative of practicing medicine, and specifically radiology, at a distance, as well as
237
providing prompt service to physician customers for use in decision making. The bottom line in
medicine is ‗‗more with less,‘‘ that is fewer personnel resources and in less time. This and a
‗‗cradle to grave‘‘ medical record is something that cannot be accomplished with film.
Digital technology is currently more expensive compared with its analog counterpart. Digital
radiography only makes economic sense at this time for dedicated chest units. Even tabletop
computed radiography readers are about twice the price of a film digitizer. Again, the limiting
factor for small clinics using radiography is personnel—lack of expertise in radiography and
short staffing. Lab techs trained in radiographic procedures are currently performing radiography
at our primary care clinics. The wider dynamic range and image processing inherent in computed
radiography cures many technical errors. Furthermore, there is no need for a film processor that
requires continuous flushing and contributes to toxic waste.
Against the proposition: Frank N. Ranallo
Opening Statement
If technological advances continue at the present pace and the world progresses economically,
one could reasonably argue that the use of film in medical imaging will fade in the far future.
But, let‘s look at the possibilities for the foreseeable future, the next 10–20 years. To try to
predict the future, it is best to first look at the past.
In 1980 Capp predicted that: ‗‗Film will be eliminated by 2000 A.D.‘‘1 In 1985 this statement
was enhanced: ‗‗A few years ago, Capp predicted that all radiology departments would be
electronic by the year 2000. A conservative estimate in 1980, it now appears that this change will
occur at least 5–10 years earlier than predicted.‘‘2 Obviously this has not occurred. In fact, sales
of film for medical imaging continue to grow, by about 5% per year in the U.S. While the use of
film for screen-film imaging has remained flat, the use of laser film has significantly increased.
The boost in laser film use over the last decade has been due to an increased number of
inherently digital imaging procedures, principally MRI and CT. This growth in laser film use
will continue in the near future with the greater use of digital receptors in conventional
radiography.
Past experience also teaches us that clear technological advances do not always totally replace
their predecessors, or at least may take quite some time to do so. Radio still exists in spite of TV,
audiocassettes are still sold alongside CDs, and I still use pencil and paper in spite of my
dependence on my computer. Economics, system performance, convenience, and ease of the
human interface usually determine the fate of new technologies. We may be approaching a phase
transition in the next 5 years that will allow digital and filmless medical imaging to grow
substantially. But will the use of film actually vanish? I think not in the foreseeable future.
CT is a totally digital acquisition technology. From its inception one has always been able to
read images from a monitor, with no loss of image quality, and with the advantage of being able
to adjust image contrast and brightness. Yet, over these many years the standard for interpreting
CT images has been to reproduce the images on film, at fixed contrast and brightness, for
viewing on illuminators. The ability to view many images together, at full resolution and with
manual random access, has been a great strength of film, since there is no need to zoom or page
through a set of images. The acceptance of soft copy diagnosis will need to be prodded by the
latest improved technology and by the enticements that its unique methods of image
management and manipulation can offer.
238
Finally the questions of image quality and economics need to be addressed. Strictly looking at
MTF and DQE curves, present CR systems offer image quality approaching but not yet equal to
screen-film. However CR‘s other advantages, such as enormous dynamic range and ability to
digitally process the image to better visualize useful clinical information, make CR quite
competitive to screen-film in clinical image quality. Flat panel (DR) technologies now in
development may even surpass screen-film in diagnostic quality. However, DR systems costing
hundreds of thousands of dollars per image receptor will find limited initial use. At present, CR
can be more economically feasible than conventional film technology, particularly for larger
departments, by eliminating the film cost, film viewers, and film processor upkeep, even though
CR cassettes and processors are substantially more expensive than their screen-film counterparts.
This initial cost, along with the additional costs of fully implementing a totally filmless
department employing PACS, may be prohibitive for smaller facilities, for some time to come.
This is particularly true for facilities with low throughput or in depressed economic areas.
Rebuttal
I agree with most of Dr. Stewart‘s arguments. We both agree that the use of filmless digital
technology will see a rapid expansion in the next few years. Our major differences may simply
involve the time frames we have chosen. I still believe that in 10–20 years the use of film will be
substantially reduced but not eliminated. Larger facilities certainly have substantial incentives to
go filmless, not the least of which is the convenience of image access to the medical staff and the
taming of fileroom problems. However, smaller clinics and medical facilities in less
economically and technologically advanced parts of the world will continue to use film in the
foreseeable future. For these latter facilities, filmless medical imaging in its present technological
state even with advances anticipated for the near future, will not offer enough incentive for
change in the next 10–20 years. For them film will remain more economically feasible while
continuing to provide quite adequate image quality and user convenience.
REFERENCES
1. M. P. Capp, ‗‗Radiological Imaging—2000 A.D.,‘‘ Radiology 138, 541–550 (1981).

2. M. P. Capp, H. R. Roehrig, G. W. Seely, H. D. Fisher, and T. W. Ovitt, Radiol. Clinics
N. Amer. 23, 349–355 (1985).
239
5.9. Lossy compression should not be used in certain imaging

applications such as chest radiography
E. Russell Ritenour and Andrew D. A. Maidment

Reproduced from Medical Physics, Vol. 26, No. 9, pp. 1773-1775, September 1999
OVERVIEW
Computational techniques are frequently used to compress image data so that transmission and
storage requirements are reduced. If the computational techniques result in no loss in image
resolution, the technique is referred to as lossless compression. Greater compression of data may
yield some loss in spatial or temporal resolution, and is referred to as lossy compression. In some
radiologic examinations [e.g., gastrointestinal (GI) studies], some resolution loss may be
tolerable, whereas in others (chest examinations and mammography) it conceivably could result
in missed pathology. Without lossy compression, however, data requirements can be
overwhelming for transmission, storage and retrieval of images such as chest films. The
unanswered question, addressed in this Point/Counterpoint issue, is whether some degree of
lossy compression can be tolerated in chest radiography.
Arguing for the Proposition is E. Russell Ritenour. Dr. Ritenour has been at the University
of Minnesota since 1989. He is Professor and Chief of Physics, Department of Radiology,
Medical School and Director of Graduate Studies in Biophysical Sciences and Medical Physics
in the Graduate School. Receiving his Ph.D. in physics from the University of Virginia in 1980,
he completed an NIH postdoctoral fellowship in medical physics at the University of Colorado
Health Sciences Center, where he remained for nine years and served as Director of Graduate
Studies in Medical Physics from 1984 to 1989. His research interests include radiologic quality
assurance, distance learning systems, and computer-based instruction.
Arguing against the Proposition is Andrew D. A. Maidment. Dr. Maidment received his Ph.D. in
Medical Biophysics from the University of Toronto in 1993. He is currently Assistant Professor
of Radiology and Director of Radiological Imaging Physics at Thomas Jefferson University in
Philadelphia. He has authored more than 65 peer-reviewed journal articles, proceedings papers
and abstracts. He has won several awards, including First Place in the 1994 Young Investigators
Competition of the International Union for Physical and Engineering Sciences in Medicine. He is
active in the ACR and AAPM, including chairing Diagnostic Imaging TG 16, Standards for
Noise Power Spectrum Analysis. His research interests include digital mammography, 3D
imaging of the breast, digital radiography detector physics and PACS.
FOR THE PROPOSITION: E. Russell Ritenour, Ph.D.
Opening Statement
At present, the only published medical standard for image quality in the realm of digital image
transmission is the ACR Standard for Teleradiology. 1 It states that ‗‗When a teleradiology
system is used to produce the official interpretation, there should not be a significant loss of
spatial or contrast resolution from image acquisition through transmission to final image
240
display.‘‘ The phrase ‗‗significant loss‘‘ is sufficiently vague that, until recognized standards-
setting organizations, such as the AMA, DICOM, FDA, or the ACR, provide specific guidance
in this area, I argue that, for legal reasons, the use of lossy compression is not advisable.
Malpractice cases require both sides to present their evidence in a way that a nontechnical
individual can understand. The outcome of this process is particularly difficult to predict when
the technology in question (and its accompanying literature) is still at a relatively early stage of
development. I maintain that the literature in the area of medical efficacy of the effects of lossy
compression is at an early stage. At such an early stage, both sides of a case may be able to use
credible expert witnesses to construct convincing cases because individual studies may,
legitimately, produce diametrically opposite results. The reasons for disagreement, which include
insufficient statistical power, the presence of confounding factors, the difference between
correlation and causation, are notoriously difficult to explain to a lay audience.
Some of the issues that relate particularly to the subject of image quality in radiology PACS and
teleradiology are also difficult to communicate to a jury of lay people. These issues include
differences in image quality among different modalities, the role of display systems, patient data
management, and communications infrastructure in the delivery of patient care, and the rapidly
evolving technology used in digital storage and transmission. Not the least among these is the
difficulty of answering the deceptively simple question: At what point in the imaging chain is the
‗‗original‘‘ image ‗‗acquired‘‘?
As the medical literature on a new technology reaches a higher level of maturity, key issues are
recognized and the criteria by which implementation of the new technology will be deemed
successful are identified. Multicenter clinical trials often appear at this stage, although not in all
cases. When the literature has reached such a level of maturity, it is possible to recognize
consensus through the appearance of the reports from government advisory groups, and
academic and professional societies.
In summary, I maintain that the use of lossy compression for some imaging procedures is
inadvisable at the present time. There is no appropriately citable medical standard. The primary
reason for this lack of standards is the relatively early stage of development of medical literature
in this area. It is an appropriate time for government advisory groups, and academic and
professional societies to begin to set standards in this important area of medicine. But, it is an
inappropriate time for individual radiologists to use lossy compression in clinical practice.
Rebuttal
I agree that there is a large and continually growing body of literature showing that lossy
compression may be used without significant degradation of image quality. However, I maintain
that this issue has two components: technical and legal. A vital link in the chain of events that
leads to a medical standard is, as of this writing, missing. This link is the presence of citable
reviews and recommendations from medical advisory groups. Without that link, individual
radiologists put themselves at legal risk.
I do not suggest that the medical community wait for long. On the contrary, the existence of this
issue speaks to a need for action. Fortunately, there are some developments in this area. One
example is a project of the AAPM Committee on Research and Technology Assessment. 2 This
project seeks to evaluate the effects of compression in musculoskeletal and thoracic images. The
committee plans to submit the results of the study to the ACR specifically for the purpose of
241
extending the current ACR Standard on Teleradiology. As of this writing, NIH funding for this
project is being sought. A corollary development is the possibility that a forthcoming revision of
the JPEG standard will include wavelet compression, one of the most successful methods of
achieving ‗‗substantial equivalence‘‘ with a high compression ratio. If so, it will probably be
adopted by DICOM since they already support JPEG. This would help to standardize procedures
and specifications.
Compression standards should be and will be adopted. However, until they are, I will continue to
advise the physicians with whom I work to avoid the use of lossy compression in images used
for primary diagnosis, particularly in imaging applications such as chest radiography.
AGAINST THE PROPOSITION: Andrew D. A. Maidment, Ph.D.
Opening Statement
Lossy compression (LC) is an indispensable part of medical imaging. The need for LC is clear—
image sizes exceed the practical and economic limits of telecommunications and storage devices.
Moreover, the initial fears that LC would mask subtle pathology have proven to be unfounded.
Study after study is showing that all imaging applications should be considered as candidates for
LC, albeit with potentially different techniques and compression ratios.
There is a definite need for LC in the transmission and storage of medical images. A typical
radiographic study will be between 20 and 100 MB. If it is necessary to send such data in a
timely fashion (i.e., a few minutes), either expensive high speed networks are required (e.g., T1
or faster), or LC must be utilized. When one considers storage of these images for five or more
years, even small institutions performing 10 000 cases per year can quickly accumulate multiple
terabytes of data. LC reduces both storage hardware and media costs, while speeding retrieval
since more image data can remain on fast devices longer.
The most common concern is that LC may suppress relevant details or inject spurious noise into
images. Such concerns are largely unfounded. The effect of LC depends upon the compression
ratio and method. As the compression ratio is increased, the first noticeable effect is the removal
of high frequency decorrelated noise, followed by increased blurring and finally by the
introduction of artifacts.3 Detectability degradation from LC can therefore be treated as being
equivalent to SNR reductions from other sources. Zhao et al.4 have shown that detectability is
equivalent for 4.5:1 LC images and uncompressed images. They have also shown that 17.4:1 LC
images are equivalent to uncompressed images, if the LC images are acquired with 25% higher
input SNR. Stated another way, a 200 speed LC screen-film image compressed 17:1 would be
equivalent to an uncompressed 300 speed screen-film image. Numerous studies have shown that
LC can even improve image quality. For example, JPEG LC has been shown to reduce speckle in
ultrasound images. LC has also been shown to offer improvement in the detection of lesions in
chest radiographs.5
Medico-legally, LC is little different from other forms of image processing. All digital imaging
modalities perform image or data processing prior to display. We accept such manipulations
through articles of faith and the presumption that FDA approval is an endorsement of the
efficacy of the device. In fact, the ACR and FDA both allow LC; they only require that the use of
LC and the compression ratio be noted on compressed images. The ACR also suggests that the
242
compression ratio be user selectable. Trained observers can thus learn to recognize compression
artifacts just as they do gridlines or processor artifacts, and compensate for them appropriately.
In summary, not all forms of LC are equal. Some will be better suited to one type of image than
another. Moreover, scientific studies must be performed prior to acceptance of specific LC uses.
However, there has been sufficient proof in the literature over a sufficiently broad range of
applications6 to demonstrate the universal acceptability of LC.
Rebuttal
It most certainly is not ‗‗an inappropriate time‘‘ to begin use of lossy compression (LC). I agree
that LC lacks an authoritative standard, but there is a DICOM Working Group addressing this
exact issue. It is important to realize that radiographic interpretation of LC images occurs daily
throughout the world. It is through LC that university medical centers can provide subspecialty
radiological expertise to small rural communities that otherwise would be served by people who
may not be adequately qualified. Is there a greater loss of information in LC of images or in the
unskilled interpretation of images?
In spite of many jokes to the contrary, people should not live their lives in fear of lawyers.
Rather, lawyers and the law can be seen to serve a constructive purpose in society. They require
us to consider the consequences of our actions. One might argue that in spite of concerns of
potential future lawsuits, medical science and in fact all fields of human endeavor continue to
develop and grow. However, it is equally possible that legal and ethical accountability
subconsciously drives us to continuously improve our existence. Such improvements necessarily
take into account societal needs, and the cost that society is willing to pay for such
improvements. LC is just one of the many improvements that allows us to implement digital
imaging in radiology with the concomitant improvements in image quality, medical care, and the
accessibility to such care. I would argue, therefore, that timely adoption of LC in radiology is a
priority so long as the conditions that require LC exist. However, as with all innovations in
radiology, LC must be properly utilized. Thus, it is important that we educate users of the uses
and potential abuses of lossy compression.
REFERENCES
1. ACR Standards for teleradiology: Diagnostic radiology standard No. 12. Reston, VA,
American College of Radiology, Rev. 26, 1996.
2. Perry Sprawls, Chair, AAPM Committee on Research and Technology Assessment
(personal communication).
3. K. Pearsons, P. Palisson, A. Manduca, B. J. Erickson, and V. Savcenko, ‗‗An analytical
look at the effects of compression on medical images,‘‘ J. Digital Imaging 10, 60–66 (1997).
4. B. Zhao, L. H. Schwarz, and P. K. Kijewski, ‗‗Effects of lossy compression on lesion
detection: Predictions of the nonprewhitening matched filter,‘‘ Med. Phys. 25, 1621–1624
(1998).
5. V. Savcenko et al., ‗‗Detection of subtle abnormalities on chest radiographs after
irreversible compression,‘‘ Radiology 206, 609–616 (1998).
6. Understanding Compression, edited by P. Drew (SCAR, Reston, VA, 1997).
243
CHAPTER 6
Mammography
6.1. All mammograms should be double-read
Craig Beam and R. Edward Hendrick

Reproduced from Medical Physics, Vol. 26, No. 2, pp. 115-118, February, 1999
OVERVIEW
Several studies have demonstrated that independent double reading of screening mammograms
improves the detection of possible cancers. However, several arguments have been made against
double reading, including: (1) a single proficient mammographer should be able to identify
virtually all detectable cancers; (2) more false positives and therefore more unneccssary biopsies
will be created by double reading; and (3) double reading increases the time but not the
reimbursement for mammography. In October 1997, the National Cancer Institute held a two-day
workshop which concluded that double reading increases the sensitivity of mammography
screening by about 5%. Still, very few centers practice double reading. This Point/Counterpoint
series explores the two sides of the issue of double reading of mammograms.
Arguing for the Proposition is Craig Beam. Craig Beam, Ph.D., is Director of Biostatistics,
Robert H. Lurie Cancer Center, Northwestern University Medical School. He is also Associate
Professor, Department of Preventive Medicine, Northwestern University Medical School. Dr.
Beam received the Ph.D. in Statistics from lowa State University in 1989. He was Assistant
Professor, Department of Radiology at Duke University from 1989 to 1992. Dr. Beam‘s research
interests are in the evaluation of diagnostic technologies. He has been working for the past 5
years developing methods for the assessment of diagnosticians. He currently is principle
investigator on an NCI-funded study of the variability in screening mammogram interpretation.
Arguing against the Proposition is Ed Hendrick. Dr. Ed Hendrick is Professor and Chief of the
Division of Radiological Sciences at the University of Colorado Health Sciences Center. He
helped develop the ACR Accreditation Programs in mammography, stereotactic imaging, and
MRI, and served as Chairman of the ACR MAP Physics Subcommittee from 1987 to 1996. He is
also Chairman of the ACR Committee on Mammography Quality Assurance, which wrote the
ACR Mammography QC Manuals. He has served as Treasurer and President of the Society for
Magnetic Resonance Imaging, as co-chair of the Agency for Health Care Policy and Research
Panel on Quality Determinants of Mammography, and as a member of the National
Mammography Quality Assurance Advisory Committee. He is a fellow of SMRI, SBI, ACR, and
AAPM.
244
For the proposition: Craig Beam
Opening Statement
The benefits and costs of double reading in radiology has been a much published topic and
various forms have been considered. Yerushalmy1 investigated the use of double reading in mass
radiology. In 1952 Groth-Peterson2 investigated the magnitude and impact of individual variation
in the reading of photoflurograms and empirically investigated the value of dual reading. A
consequence of this study was that dual reading was implemented in a mass screening campaign
against tuberculosis in Denmark. Rimington3 investigated two forms of double reading (in one
form only negative films were initially reread) and concluded that neither form of double reading
ought to be implemented in the mass screening of chest x rays because the administrative costs
outweighed the slight increase in active cases. A 1976 study4 of the value of consultation among
radiologists found consensual diagnosis to outperform simple majority rule based on independent
interpretations. Stitik5 concluded that the economic burden from double reading might be
reduced if nonphysicians could do screening. Hessel6 investigated the effectiveness and cost of
an independent third arbiter in chest radiograph interpretation and concluded that ‗‗the choice
between single and multiple interpretations must be evaluated in each clinical setting and should
consider expected improvements in accuracy, implications for patient care, and additional costs
(p. 589).‘‘ Anttinen7 and Thurfjell8 investigated the use of independent double reading in
screening mammography. Both studied two radiologists and found double reading to increase
case detection. However, although Thurfjell‘s method was limited to the use of independent
interpretations, Antinnen‘s study included a consultation between the two radiologists whenever
a case was considered for recall by one of the radiologists. A recent study of the impact of reader
variability in double reading demonstrated that the effectiveness of double reading can vary
widely depending on the reading pair.9
In reality, double reading has many possible forms, each of which need to be objectively
evaluated for their impact to sensitivity, false positive errors, and practicality. Double reading in
screening mammography should be treated like other diagnostic adjuncts, such as computer-
assisted diagnosis, and developed as a useful technology in its own right.
Rebuttal
Dr. Hendrick raises several important issues related to the use of double reading in screening
mammography. Among these are the increased rate of false positives, the increased costs
required to implement double reading, and problems associated with the practical
implementation of double reading. Of course, none of these problems is unique to double
reading, they beset almost every technological innovation in radiology.
Dr. Hendrick cites research I have conducted as an argument against double reading 9 (see also
Ref. 10). The results we found in doing this research speak, on average, against widespread
adoption of independent double reading for the very reasons pointed out by Dr. Hendrick.
However, the study analyzed the extent of variability in the effectiveness of double reading, as
well as the average performance in a population of radiologists. We found that the impact of
double reading can be highly variable, and possibly advantageous for some pairs of readers. An
example given in our paper shows a distinct improvement in the sensitivity of one of two
radiologists with a very minimal increase in false positives.
245
The key point to be underscored is that double reading needs to be evaluated in the many forms
that it might take. Furthermore, since human variability can exercise a strong impact on its
effectiveness, double reading might be best evaluated at the level of individual practices, leaving
the decision whether or not to implement this adjunct up to radiologists.
REFERENCES
1. J. Yerushalmy, J. T. Harkness, J. H. Cope, and B. R. Kennedy, ‗‗The role of dual reading

in mass radiography,‘‘ Am. Rev. Tuberculosis 61, 443– 464 (1950).
2. E. Groth-Peterson, A. Lovgreen, and J. Thillemann, ‗‗On the reliability of the reading of
photofluorograms and the value of dual reading,‘‘ Acta Tuber Scandia 26, 13–37 (1952).
3. J. Rimington, ‗‗Double reading by the same reader in mass radiography,‘‘ Tubercle 48,
38–44 (1967).
4. B. J. Hillman, R. G. Swensson, S. J. Hessel, D. E. Gerson, and P. G. Herman, ‗‗The value
of consultation among radiologists,‘‘ Am. J. Roentgenol., Radium Ther. Nucl. Med. 127, 807–
809 (1976).
5. F. P. Stitik and M. S. Tockman, ‗‗Radiographic screening in the early detection of lung
cancer,‘‘ Radiologic Clinics of North America 15, 347– 366 (1978).
6. S. J. Hessel, P. G. Herman, and R. G. Swensson, ‗‗Improving performance by multiple
interpretations of chest radiographs: effectiveness and cost,‘‘ Radiology 127, 589–594 (1978).
7. I. Anttinen, M. Pamilo, M. Soiva, and M. Roiha, ‗‗Double reading of mammography
screening films—one radiologist or two?‘‘ Clinical Radiol. 48, 414–421 (1993).
8. E. L. Thurfjell, K. A. Lerneval, and A. A. S. Taube, ‗‗Benefit of independent double
reading in a population-based mammography screening program,‘‘ Radiology 191, 241–244
(1994).
9. C. A. Beam, D. C. Sullivan, and P. M. Layde, ‗‗Effect of human variability on
independent double reading in screening mammography,‘‘ Academic Radiology 3, 891–897
(1996).
10. C. A. Beam and D. C. Sullivan, ‗‗What are the issues in the double reading of
mammograms?‘‘ Radiology 193, 582 (1994).
Against the proposition: R. Edward Hendrick
Opening Statement
Double reading of mammograms is not common among U.S. mammography practices.

Requirement of double reading at U.S. mammography practices would have marginal benefit,
would trigger an excessive number of negative biopsies, and would be financially disastrous.
The asserted benefit of independent double reading, an increased sensitivity of approximately

5%, is achieved at the expense of 15% higher call-back rate.1 Dr. Beam‘s own research showed
that the improvement in sensitivity due to double reading is highly variable and depends on the
particular pair of radiologists involved.2 His research also provided the methods to show that in a
screening population where the incidence of breast cancer is 3–6 per 1000, as in the typical U.S.
mammography practice,3 double reading would cause the average radiologist to trigger 167 to
335 additional negative biopsies for each additional cancer found. 2 Thus, double reading by the
average U.S. radiologist pair would have a small, highly variable benefit, at tremendous added
cost, both economic and psychological, due to the large number of negative biopsies that double
reading causes.
246
Current Medicare reimbursement for screening mammography is $64.73, $44.02 for technical
and $20.71 for interpretation fees. 4 At current reimbursement rates, both screening and
diagnostic mammography lose money at most mammography sites. MQSA Final Rules going
into effect in 1999 will add several dollars to the cost of each woman‘s mammogram due to
added equipment and direct patient notification of results requirements. 5 Requiring double
reading of mammograms, whether screening or diagnostic, would add logistic difficulties and
unreasonable costs for mammography practices.
Diagnostic mammography already requires a radiologist to be on site during the procedure.

Double reading of diagnostic mammograms in a timely manner that might affect the diagnostic
decision would require two radiologists to be on site and available to interpret mammograms, a
practical impossibility at most sites.
Most screening mammograms are batch-read and would appear to better accommodate double
reading. There is still the need, however, to interpret screening exams and to deliver
mammography reports in a timely fashion to patients and their referring physicians. Double-
reading screening mammograms would require bringing two radiologists to the same place to
batch-read films placed on the same alternator or would require delivering films to two different
sites for loading on two different alternators. The first alternative might work at some larger
practices, but would be impractical at the great majority of smaller, private mammography
practices. The second alternative is time-consuming and expensive for any practice. In addition,
the standard practice of marking the locations of suspicious findings on films for follow-up or
biopsy would have to be eliminated to permit independent double reading.
Who will pay the added costs of a second interpretation, additional film handling, and more
complicated reporting caused by double reading? To ask U.S. mammography practices to absorb
these added costs within the current reimbursement scheme would spell economic disaster for
most sites. To expect Medicare, insurance companies, and women themselves to pay an
additional $25–30 per screening mammogram to cover the costs of double reading is unrealistic.
When these payers learn that double reading yields a marginal and variable increase in
sensitivity, at the price of a tremendous increase in negative biopsies and their costs, their
support for double reading is likely to vanish.
Rebuttal
I agree with Dr. Beam that double reading has many forms, each of which needs to be evaluated
for its impact on sensitivity, false positives, and clinical practicality. This is far short of agreeing
that double reading is ready for implementation in clinical practice.
Currently, direct digital acquisition and computer-aided detection methods are in the process of
receiving FDA approval as clinical tools in mammography, with improved methods for
acquisition and detection evolving rapidly. This is exactly the wrong time to presume that double
reading by human observers is the best or most efficient interpretation scheme to improve
mammography outcomes.
Double reading may have some role as a quality assurance measure in mammography, if results
of double reading can be used as an effective quality improvement tool. That is far short,
however, of accepting double reading as a standard in clinical practice.
247
The best argument against the clinical effectiveness of double reading comes from Dr. Beam‘s
1996 paper on the subject:6 ‗‗Our analyses also show that, because of human variability,
independent double reading in screening mammography can have a widely variable effect. Not
all radiologists complement each other diagnostically, and some add little, if anything, to the
accuracy of the other. Adding the reading from a more experienced radiologist does not
necessarily improve the TPR [true positive rate or sensitivity] of a radiologist with less
experience.‘‘
As Dr. Beam suggests in his opening statement, various forms of double reading need to be
compared to single reading and to different methods of reading with the aid of computer-aided
detection methods, preferably in a realistic patient population. This should be done for both film
and digital mammography before we can conclude that one interpretation scheme is best for
clinical practice. Such assessments should include cost-benefit analyses that compare not only
the costs and benefits of the interpretation scheme, but also the costs and benefits of the
additional procedures triggered by the specific double reading method. Only if cost-benefit
analyses prove favorable for double reading should it be considered as a standard of clinical
practice for screening or diagnostic mammography. Then double reading might become one of
those rare areas in medicine where appropriate research preceded adoption into clinical practice.
REFERENCES
1. S. Ciatto et al., ‗‗Independent double reading of screening mammograms,‘‘ J. Med.

Screening 2, 99–101 (1995).
(1996).
3. M. N. Linver and S. B. Paster, ‗‗Mammography outcomes in a practice setting by age:
prognostic factors, sensitivity, and positive biopsy rate,‘‘ Monogr. Natl Cancer Inst. 22, 113–117
(1997).
4. American College of Radiology (private communication).
5. 21 CFR Parts 16 and 900: Quality Mammography Standards; Final Rules. Federal
Register, October 28, 1997, Vol. 62: No. 208, 55852–55994.
(1996).
248
6.2. Pre-menopausal women should be actively encouraged to seek

screening mammograms
Andrew Maidment and Elizabeth Krupinski

OVERVIEW
The controversy continues over the merits of screening mammography for pre-menopausal women
(women<50 years of age). Many persons believe that screening benefits are intuitively obvious, and are
well supported by experimental evidence. These individuals endorse screening mammograms for
younger women. Others feel that the scientific evidence for screening mammography is inconclusive,
and that this uncertainty should be communicated to younger women contemplating mammography.
Until now, medical physicists have been relatively silent on the screening issue. But this
Point/Counterpoint breaks the silence.
Arguing for the Proposition is Andrew Maidment, Ph.D. Dr. Maidment received his Ph.D. in Medical
Biophysics from the University of Toronto in 1993. He is currently Assistant Professor of Radiology,
and Chief, Physics Section at the University of Pennsylvania in Philadelphia. He has more than 110 peer-
reviewed journal articles, book chapters, proceedings papers and abstracts. He has won several awards
including First Place in the 1994 Young Investigators Competition of the International Union for
Physical and Engineering Sciences in Medicine. He is active in the ACR and AAPM, including chairing
Diagnostic Imaging TG 16, Standards for Noise Power Spectrum Analysis. His research interests include
digital mammography, 3-D x ray imaging of the breast, and digital radiography detector physics.
Arguing against the Proposition is Elizabeth Krupinski, Ph.D. Dr. Krupinski received her undergraduate
education at Cornell and her Ph.D. at Temple University, both in Experimental Psychology. She has
been at the University of Arizona since 1992 in the Departments of Radiology and Psychology. Her
interests lie in medical image perception and decision-making, especially in the digital environment. The
human-computer interaction is also of interest from the human factors perspective. She is interested in
the causes of interpretation error and in developing ways to improve training from an image perception
perspective. Dr. Krupinski is also the Associate Director of Evaluation for the Arizona Telemedicine
Program.
FOR THE PROPOSITION: Andrew Maidment, Ph.D.
Opening Statement
There is almost complete consensus that routine mammographic screening can reduce the
mortality of breast cancer. Recent results indicate mortality can be reduced by 40 to 45%.1
Controversy continues, however, over whether this reduction is shared by all women or whether
it begins after menopause, at approximately 50 years of age in developed countries. This
dichotomistic doctrine is fallacious; all women should be actively encouraged to seek screening
mammograms starting at age 40.
249
There is little that distinguishes breast cancer in a woman in her 40s from that in a woman in her
50s. The natural incidence increases only slightly between the two decades. 2 The etiology,
pathology and clinical sequelae are virtually identical. Furthermore, while women 40–49 account
for only 16% of breast cancer incidence, they account for 40% of the years of life lost to breast
cancer.3 Thus, women in their 40s will potentially benefit most from screening. The meta-
analysis of randomized clinical trials (RCTs) by Humphrey et al.4 indicates that the summary
relative risk for women of ages 40–49 is 0.80 (CI 0.67–0.96), compared with a summary relative
risk of 0.78 (CI 0.70–0.87) for women older than 50. These estimates correspond to one life
saved per 1385 women for the younger group and one life saved per 838 for the older group.4
Thus, the benefit of routine screening for both groups of women is comparable.
Admittedly, there is not universal agreement on this issue. Notably, the Cochrane Report5 found
no benefit for women aged 40–49. However, the Cochrane Report considered only 2 of 8
applicable RCTs. One of these, the CNBSS study, has been the subject of extensive criticism.6
That said, mammography is far from perfect. Mammography lacks sensitivity; some cancers are
missed. Mammography also lacks specificity; many healthy women endure negative biopsies,
resulting in a high monetary cost, as well as physical and psychological costs. Mammography
entails the risk of inducing cancers, but this risk must be weighed against the probable benefits. It
is estimated that annual screening from age 40 to 49 will induce fewer than 8 cancers per 100,000
women screened.7 Thus, nine lives would be saved for each woman who suffers an iatrogenic
cancer. Mammography may find indolent cancers. Finally, universal screening guidelines are
questionable; for example, consider women with genetic predisposition to cancer (e.g., BRCA-
1/2). These are not failings of mammography alone, however; they demonstrate the boundaries of
our knowledge of breast cancer, and limitations of current diagnostic and treatment methods.
Mammography may have some flaws, but it is the best screening tool for breast cancer available
today. Above the age of 50, it is almost universally accepted that the death rate from breast
cancer can be reduced at a monetary, physical and psychological cost that society accepts. There
is no evidence to support a different approach for women in their 40s. Finally, early
mammograms provide a highly valuable baseline for radiologists attempting to interpret
mammograms later in a woman's life. Thus, a variety of considerations strongly support
extension of the benefits of routine mammographic screening to this younger population. That is,
all women should be actively encouraged to seek screening mammograms starting at age 40.
Rebuttal
Even casual readers of both my own and my colleague's opening statements on this topic will
remember Leonard Courtney's famous words "lies — damn lies — and statistics." 8 First, I
disagree with the assertion that mammography and menopause should be tightly linked. The
mean age of menopause in the US is 51, not 45.9 Menopause is exceptionally rare below the age
of 40 or above the age of 59.9 If menopausal changes had the dominant responsibility for breast
cancer incidence, then there should be a stronger correlation with menopause. Yet the strongest
correlate of breast cancer risk is age, increasing almost linearly with age from 30 to 75.9
Second, the data related to positive-predictive value (PPV) cited by my colleague in this
Point/Counterpoint are dated and fail to distinguish prevalence (first) screening from subsequent
screenings. Consider for example the Ghent program; PPV = 14.2% for prevalence screens of
women in their 40s, while PPV = 28.3% for prevalence screens of women in their 50s. However,
250
PPV = 19.7% and 16.8% in subsequent screens in 40s and 50s, respectively. The increased PPV
of the 50s prevalence screen is due to the fact that screening has been started too late for these
women; they already have a significant number of readily-detectable cancers. Moreover, the
equal values of PPV in subsequent screenings in both age groups clearly indicate that
mammography is equally effective for both.
The assertion that young dense-breasted women benefit less from mammography is also
questionable. Kerlikowske10 has shown that sensitivity in women aged 50 and older is affected by
breast density (98.4% fatty vs 83.8% dense; P<0.01), yet for women younger than 50 this is not
true (81.8% fatty vs 85.4% dense).
Screening will benefit from advances in breast cancer biology, better diagnostic tools and
improved treatments. There is little doubt, however, that mammography for women aged 40–49
is not only appropriate, but essential. Likewise, clear, consistent and simple screening guidelines
are essential. "Start annual screening on your 40th birthday" fulfills this role exactly; it is the
perfect birthday gift for any woman.
AGAINST THE PROPOSITION: Elizabeth Krupinski, Ph.D.
Opening Statement
Although the benefits of mammography for early detection and treatment of breast cancer may
seem obvious, there is still considerable debate regarding its overall efficacy, who should be
screened and at what age.11,12 The majority of trials (with findings both for and against screening)
have been done with women 40 years of age and older. Although the general consensus is that
screening mammography is useful for women over 40 (and more so as women get older), the
evidence regarding benefits for women under 40 is scarce. Menopause typically occurs between
45 and 50 years of age, with the last two years of perimenopause starting the accelerated decline
in estrogen levels. The incidence of breast cancer, to a large extent, parallels menopause onset.
Incidence is very low for women in their twenties, increases gradually and plateaus at 45, then
increases dramatically after 50. In fact, approximately 50% of breast cancers are diagnosed in
women over 65, and recent evidence indicates that since the 1980s breast cancer incidence rates
have increased only in women over 50.12,13 Invasive breast cancer diagnoses in women over 65
accounts for 45% of all new breast cancer cases, and 45% of all breast cancer deaths are in
women over 65.12
In terms of sensitivity and specificity, screening mammography is less effective in women with
dense breasts, especially younger women.14,15 The positive predictive value (PPV) ranges from
20% in women under 50 to 60% to 80% in women over fifty.16,17 The low PPV reflects the
higher false positive rate17 for younger women. Although the psychological effects of false
positives are generally short-lived and have few lasting consequences,12 the immediate
experience produces high levels of anxiety, especially since waiting times can be prolonged
between initial report and follow-up procedures. In addition to low sensitivity, specificity and
PPV in younger women, repeated screening exams starting at a younger age lead to an increased
risk of radiation-induced breast cancer.12,18 This is especially true for women with a family
history (i.e., genetic predisposition) of breast cancer, or for women being treated with radiation
for other purposes (e.g., radiation treatment, scoliosis progression imaging).
251
The relative lack of efficacy, and the potential for physical and psychological risks, support the
proposition that most premenopausal women should not be encouraged to seek screening
mammograms. Younger women who are at risk because of a family history or known genetic
predisposition to breast cancer (5–10% of all cancers) should be screened, because their
cumulative risk of breast cancer is higher than average.12 For the average premenopausal woman,
a careful analysis of the risk factors associated with breast cancer, and adherence to a healthy
lifestyle based on prevention, may be more useful than screening mammography. Although
certain risk factors cannot be altered (e.g., age of first menarche, late menopause), there are many
others that can be controlled, such as not smoking, having children early in life, increased
physical activity, maintaining proper weight, reduced alcohol intake, breastfeeding rather than
bottle-feeding, and sticking to a healthy diet.19 Educating women about these risk-reduction
factors, and suggesting other methods of screening that do not involve radiation exposure (e.g.,
ultrasound, MRI), should be the focus of communication to younger women contemplating the
costs and benefits of mammography.
Rebuttal
"Statistics—the only science that enables experts using the same figures to draw different
conclusions." 20 Reading the statistics on breast cancer screening often leads one to this very
conclusion. The Humphrey et al.21 report does indeed report that the relative risk for women aged
40–49 is 0.80, and 0.78 for women older than 50, based on their meta-analysis of eight high
profile breast screening trials. This report brings up several other points, however, that lead one
to question the strength and generalizability of the conclusions. Most important is the authors'
rating of the quality of original studies used in the meta-analysis. Each of the screening trials
included in the meta-analysis had important methodological flaws, and seven of the eight studies
were rated only fair in terms of study quality. The eighth was rated poor. Also, the authors state
that of the seven trials conducted since 1963 that included women aged 40–49, only one actually
planned to evaluate breast cancer screening in this group, and none (even the one that specifically
included it as a statistical variable) had sufficient statistical power. The lack of power is due
mainly to inadequate sample size once data were stratified into age subgroups. There is also some
question22 about when the benefits of screening mammography actually appear in the 40–49 age
group in these trials. The potential survival benefit in women aged 40–49 is typically not
observed until the trial has progressed for several years. The women included in the studies are
now just in the over-50 age group.
In the end, each woman must make a personal decision by trying to understand the overall
picture, including an understanding of absolute risk, relative risk and the factors that contribute to
breast cancer risk. No studies have been designed that offer guidance on how an individual
woman can assess her lifestyle, family history, and environment in the context of available
medical evidence to decide when and how often she should be screened. Clearly there are women
at higher risk, for whom this decision may be easier. But for women not at obvious risk, the use
of the single variable of age (other than gender of course) to determine when screening should
begin may not be sufficient. To improve breast cancer screening outcomes, we need to develop
better and more accurate models that include as many risk factors as possible for each individual
woman.
REFERENCES
252
1. S. W. Duffy et al., "The impact of organized mammography service screening on breast

carcinoma mortality in seven Swedish counties," Cancer 95(3), 458–469 (2002).
2. National Cancer Institute, "SEER Cancer Statistics Review 19752000: Breast,"
http://seer.cancer.gov/csr/1975_2000/results_merged/sect_04_breast.pdf (accessed July 24,
2003).
3. D. B. Kopans, "Screening mammography and the controversy concerning women aged
40–49 years," in Syllabus: A Categorical Course in Breast Imaging, edited by D. B. Kopans and
E. B. Mendelson (Radiological Society of North America, Oak Brook, 1995), pp. 39–50.
4. L. L. Humphrey, M. Helfand, B. K. S. Chan, and S. H. Woolf, "Breast cancer screening:
A summary of the evidence for the U.S. Preventive Services Task Force," Ann. Intern Med.
137(5), Part 1, E347–E367 (2002).
5. O. Olsen and P. C. Gøtzsche, "Cochrane review on screening for breast cancer with
mammography [Letter]," Lancet 358, 1340–1342 (2001).
6. N. F. Boyd et al., "A critical appraisal of the Canadian National Breast Cancer Screening
Study," Radiology 189(3), 661–663 (1993).
7. S. A. Feig and R. E. Hendrick, "Radiation risk from screening mammography of women
aged 40–49 years," JNCI Monogr. 22, 119–124 (1997).
8. L. Courtney, "To my fellow-disciples at Saratoga Springs," Nat. Rev. (London) 26, 21–
26 (1895).
9. I. Kato et al., "Prospective study of factors influencing the onset of natural menopause,"
J. Clin. Epidemiol. 45(12), 1271–1276 (1998).
10. K. Kerlikowske et al., "Effect of age, breast density, and family history on the sensitivity
of first screening mammography," JAMA, J. Am. Med. Assoc. 276(1), 33–38 (1996).
11. V. P. Jackson, "Screening mammography: controversies and headlines," Radiology 225,
323–326 (2002).
12. R. A. Smith et al., "American Cancer Society guidelines for breast cancer screening:
Update 2003," Ca-Cancer J. Clin. 53, 141–169 (2003). Cancer Facts and Figures 2003
(American Cancer Society, Atlanta, 2003). first citation in article
13. E. Ziv, J. Shepherd, R. Smith-Bindman, and K. Kerlokowske, "Mammographic breast
density and family history of breast cancer," JNCI, J. Natl. Cancer Inst. 95, 556–558 (2003).
14. H. Jonsson, L. G. Larsson, and P. Lenner, "Detection of breast cancer with
mammography in the first screening round in relation to expected incidence in different age
groups," Acta Oncol. 42, 22–29 (2003).
15. R. Ferrini, E. Mannino, E. Ramsdell, and L. Hill, "Screening mammography for breast
cancer: American College of Preventative Medicine practice policy statement," Am J Prev Med
12, 340–341 (1996).
16. K. Kerlikowske et al., "Positive predictive value of screening mammography by age and
family history of breast cancer," JAMA, J. Am. Med. Assoc. 270, 2444–2450 (1993).
17. D. A. Hill, S. Preston-Martin, R. K. Ross, and L. Bernstein, "Medical radiation, family
history of cancer, and benign breast disease in relation to breast cancer risk in young women,"
Cancer Causes Control 13, 711–718 (2002).
18. F. Lumachi et al., "Breast cancer risk in healthy and symptomatic women: results of a
multivariate analysis. A case-control study," Biomed. Pharmacother 56, 416–420 (2002).
19. E. Esar, Esar's Comic Dictionary (Doubleday, London, 1983). first citation in article
20. L. L. Humphrey, M. Helfand, B. K. S. Chan, and S. H. Wolf, "Breast cancer screening: a
summary of the evidence for the US Preventive Services Task Force," Ann. Intern Med. 137,
E347–E367 (2002).
21. Susan G. Komen Breast Cancer Foundation http://www.komen.org/bci/ (accessed August
4, 2003).
253
6.3. Computer-aided detection, in its present form, is not an effective

aid for screening mammography
Robert M. Nishikawa and Maria Kallergi

OVERVIEW
An obvious application of computer-aided detection (CAD) systems is in the interpretation of

screening mammograms. The intent is to prompt the radiologist to cancers that otherwise might
have been missed in the first reading of the mammogram. Indeed, several studies have claimed to
demonstrate that CAD is beneficial for screening mammography. For example, in a report on a
large prospective study in the US it was concluded that: "The use of CAD in the interpretation of
screening mammograms can increase the detection of early-stage malignancies."1 However,
recently published results from a similarly large study in the UK have been claimed to show that
CAD, in its present form, is not effective in that there was no significant difference observed in
cancer detection rates with CAD compared with reading mammograms without CAD.2 This claim
forms the basis of this month's Point/Counterpoint debate.
Arguing for the Proposition is Robert M. Nishikawa, Ph.D. Dr. Nishikawa received his B.Sc. in
Physics in 1981 and his M.Sc. and Ph.D. in Medical Biophysics in 1984 and 1990, respectively,
all from the University of Toronto. He is currently an Associate Professor in the Department of
Radiology and the Committee on Medical Physics at the University of Chicago. He is a Fellow of
the AAPM. His current research interests are in digital radiography, computer-aided diagnosis,
mammography, and measures of image quality.
Arguing against the Proposition is Maria Kallergi, Ph.D. Dr. Kallergi received her Ph.D. in Physics and
Electrical Engineering from the University of South Florida in 1990. Her background was in digital
detectors and semiconductor devices but, in 1991, she started a career in medical imaging as a
Postdoctoral Research Fellow at the H. Lee Moffitt Cancer Center & Research Institute and the
Department of Radiology at the University of South Florida, where she is currently an Associate
Professor of Radiology. Dr. Kallergi's research interests focus on two- and three-dimensional digital
breast imaging technologies and computer methodologies for the automated detection and diagnosis of
breast, lung, and pancreatic cancers.
FOR THE PROPOSITION: Robert Nishikawa, Ph.D.
Opening Statement
For CAD to be effective, two conditions must exist. Firstly, the CAD system must be capable of
detecting cancers that are missed clinically. Secondly, the radiologist must be able to recognize when the
computer has found a missed cancer. There is ample evidence in the literature to support the first
condition. However, published sensitivities for the detection of clinically missed breast cancers range
from 50%–80%.3,4,5 Hence, there is not conclusive evidence that the second condition is true.
254
There have been four published clinical evaluations of the use of CAD for the detection of breast
cancers.1,6,7,8 Two of these were inconclusive,6,7 one because it was only a pilot study,7 and two showed
an apparent benefit for using CAD,1,8 there was an increase in the number of cancers detected. However,
neither of these two studies was statistically significant, probably because breast cancer prevalence in a
screening population is low. In one of the two CAD-beneficial studies, clustered microcalcifications
were detected in seven of the eight missed cancers that were detected by the computer and only one was
a mass.1 In the pilot study, the only CAD-detected cancer was a cluster of microcalcifications. The two
other studies were retrospective temporal comparisons and it is not possible to determine what type of
lesions the CAD system helped the radiologists find.
CAD detection of clustered microcalcifications can be effective in helping radiologists find cancers. This
is because the sensitivity of the scheme is high—up to 98%. Further, very little normal breast anatomy
mimics calcifications in a mammogram. Therefore, radiologists can easily dismiss false detections.
However, radiologists do not miss calcifications as often as they miss masses. In two large studies, 30%
of missed cancers demonstrated calcifications, while 70% demonstrated masses. 3,4 This implies that
radiologists either do not look at the computer prompt for masses, or that they look at the prompted area
but do not recognize that a cancer is present. As shown by Zhang et al., one reason why radiologists tend
to ignore correct prompts is that the false detection rate of the CAD scheme is too high.9 The
superposition of normal breast structure often mimics a mass. Not only does this increase the false-
detection rate of the CAD scheme, but it also makes it difficult for radiologists to determine if the
computer is prompting a true or false lesion.
I believe that computer-aided diagnosis will one day be pervasive throughout radiology. However, it is
important that we understand its limitations, both technical and clinical. In this way we can improve the
technology and this will lead to faster clinical implementation and acceptance. In my opinion, further
research is needed to improve CAD for mammography, particularly for mass detection, so that it can
truly be an effective clinical tool for radiologists. CAD for screening mammography is not quite there
yet.
AGAINST THE PROPOSITION: Maria Kallergi, Ph.D.
Opening Statement
Success depends on a number of factors when assessing the effectiveness of any new medical
technology, including computer-aided detection systems for screening mammography.
Firstly, there is the question of whether CAD has fulfilled its initial goal. CAD was intended as an aid to
the radiologist. Its primary goal was to help detect breast calcifications and spiculated masses at an
earlier stage. There are now several testimonies that current CAD systems meet this challenge with a
positive impact on cancer detection rates particularly for masses. 8,10,11,12 In addition, callback rates, a
major concern initially, have not been affected13,14 and the false positive signals, a major turn-off factor
in the first generation of CAD systems, have been significantly reduced in subsequent upgrades (by
almost a factor of 2) and can now be handled with greater comfort and dismissed easily by the trained
reader.15
In terms of clinical acceptance, ten years ago, when these systems were still in the research laboratories,
the medical community was divided, with unfriendly skeptics being an overwhelming majority. Today,
radiologists are still divided but the friendly group has grown bigger and friendlier and many of the
skeptics have become converts. This change is a consequence of the clinical implementation of CAD that
255
has allowed the "learner" to practice and acquire the necessary skills for its use. It is also due to the
positive results of recent studies that evaluated the systems clinically.10 Radiologists now find that CAD
has increased their level of confidence and that its role is not only that of a "second reader" but also a
"refocusing tool" in the often monotonous and repetitive environment of screening mammography,
where external factors cause attention interruption and possible observational oversight.
CAD has also had a positive impact beyond the radiologist: on the patient and the administrator. Patients
appear more comfortable with the mammographic procedure, they demonstrate an increased level of
security, and are more amenable to radiology decisions. Administrators are more likely to embrace CAD
now because it can be seen as a marketing tool that could benefit the institution and, in some practices, it
even makes good fiscal sense due to the higher reimbursement rate of the mammography package that
includes CAD.
Admittedly, the current commercial systems have problems and limitations. Both are well known and
understood. It is also well known, but poorly understood by those outside the CAD community, that
CAD requires continued commitment and investment to reach the desired levels of performance.
Although far from perfect, today's CAD is steadily shaping its role as part of the standard of care.16 This
is due to its demonstrated success in improving (i) the sensitivity of screening mammography and the
confidence of the reader, (ii) the security and comfort of the patient, and (iii) the workflow and financial
prospects of the administrator.
Rebuttal: Robert Nishikawa, Ph.D.
My colleague, Maria Kallergi, raises a number of important points. The most important one on which we
agree is that further research is needed to improve CAD, because existing systems have problems and
limitations. I, however, disagree that these problems are all known and well understood. I think the
biggest problem is that experience has shown that radiologists frequently ignore "missed" lesions
detected by CAD that turn out to be true positives. As Birdwell et al. noted, the cancer detection rate
increased by 7.4% in their clinical study, whereas, in their retrospective study, CAD was able to detect at
least 27% of "actionable" cancers.4,11 This difference is one of the principle reasons why CAD is not
effective in screening mammography.
Further I do not agree that callback rates remain unaffected by CAD. Studies by Freer 1 (18%), Helvie7
(9.7%), Birdwell11 (8.2%) and Cupples8 (10.9%) all found that the callback rate increased when CAD
was used. Further research is needed to understand whether the increase in sensitivity justifies the
increased callback rate, although I believe that it probably does.
I do agree that there are intangible benefits of CAD to the radiologist. Higher confidence and less fatigue
are just two. However, one would hope that this would result in higher performance. This has not yet
been documented conclusively.
Defining the benefits of CAD in screening mammography is difficult because of the low cancer
prevalence in the screening population. Clinical evaluations are the only true measure of clinical
effectiveness. Retrospective studies, two of which Dr. Kallergi quotes as proof of the benefits of CAD,
are not definitive and are often overly optimistic. No clinical study has demonstrated a statistically
significant benefit for using CAD.
I believe that CAD systems will improve and become the standard of care in the future, but those CAD
systems of the future will necessarily have higher performance than the systems studied to date. Newer
versions with higher performance are now available and should be tested clinically.
256
Rebuttal: Maria Kallergi, Ph.D.
I do not disagree with my esteemed opponent on technical issues or future requirements of CAD systems
for mammography. I disagree on the way success is judged. I argue that the usefulness of CAD
technology today cannot be judged by diagnostic benefits alone. I further argue that benefits come in
small "doses," as is the case for the vast majority of medical technologies, and depend on a variety of
factors including practice type and volume, readers' experience, and method of interpretation. But even if
we focus on the diagnostic benefits alone, recent studies provide convincing evidence that the two
conditions set by Dr. Nishikawa as being necessary to demonstrate that CAD is effective, are met by
current systems. Specifically, the work of Brem et al.10 supports the first condition related to the
capabilities of CAD, and that of Birdwell et al.11 supports the second condition related to the readers'
ability to interpret the CAD results correctly. This was a large prospective study, which demonstrated the
clinical benefits of CAD with the surprising outcome that CAD was shown to be helpful in detecting
masses that had been missed by the radiologist, whereas previous studies had shown most benefit for the
detection of calcifications. These findings, along with results from previous studies, 8 strengthen the
arguments against the Proposition and there is now substantial evidence that current CAD systems are
effective clinically in more than one way.
If one problem has to be identified that potentially impacts negatively on the current CAD systems, it is
that the marketing part of the technology was pursued before the science. The result has been the
creation of a negative bias on the side of users, mainly due to high false detection rates. This bias has
been, and still is, difficult to overcome. Also, we have been clumsy and unprepared for the training
process. This has led to a longer and more difficult learning curve that has generated more bias and
skepticism. Taking these two elements into account, it is not difficult to explain the inconclusive and
sometimes contradictory results of earlier studies.
It cannot be denied that definitive diagnostic improvements will assuage all doubts regarding the
usefulness of CAD in mammography. This, however, should not be our only criterion of success. If all
factors are considered, we can conclude that current CAD systems, although not at optimum
performance, are useful. This allows us to dream of what it will be like when optimum performance is
reached.
REFERENCES
1. T. W. Freer and M. J. Ulissey, "Screening mammography with computer-aided detection:

prospective study of 12,860 patients in a community breast center," Radiology 220, 781–786 (2001).
2. L. A. L. Khoo, P. Taylor, and R. M. Given-Wilson, "Computer-aided detection in the United
Kingdom national breast screening programme: prospective study," Radiology 237, 444–449 (2005).
3. L. J. W. Burhenne, S. A. Wood, C. J. D'Orsi, S. A. Feig, D. B. Kopans, K. F. O'Shaughnessy, E.
A. Sickles, L. Tabar, C. J. Vyborny, and R. A. Castellino, "Potential contribution of computer-aided
detection to the sensitivity of screening mammography," Radiology 215, 554–562 (2000).
4. R. L. Birdwell, D. M. Ikeda, K. F. O'Shaughnessy, and E. A. Sickles, "Mammographic
characteristics of 115 missed cancers later detected with screening mammography and the potential
utility of computer-aided detection," Radiology 219, 192–202 (2001).
5. R. F. Brem, J. Baum, M. Lechner, S. Kaplan, S. Souders, L. G. Naul, and J. Hoffmeister,
"Improvement in sensitivity of screening mammography with computer-aided detection: a
multiinstitutional trial," Am. J. Roentgenol. 181, 687–693 (2003).
6. D. Gur, J. H. Sumkin, H. E. Rockette, M. Ganott, C. Hakim, L. Hardesty, W. R. Poller, R. Shah,
and L. Wallace, "Changes in breast cancer detection and mammography recall rates after the
introduction of a computer-aided detection system," J. Natl. Cancer Inst. 96, 185–90 (2004).
257
7. M. A. Helvie, L. Hadjiiski, E. Makariou, H. P. Chan, N. Petrick, B. Sahiner, S. C. Lo, M.

Freedman, D. Adler, J. Bailey, C. Blane, D. Hoff, K. Hunt, L. Joynt, K. Klein, C. Paramagul, S. K.
Patterson, and M. A. Roubidoux, "Sensitivity of noncommercial computer-aided detection system for
mammographic breast cancer detection: pilot clinical trial," Radiology 231, 208–214 (2004).
8. T. E. Cupples, J. E. Cunningham, and J. C. Reynolds, "Impact of computer-aided detection in a
regional screening mammography program," Am. J. Roentgenol. 185, 944–950 (2005).
9. B. Zheng, R. G. Swensson, S. Golla, C. M. Hakim, R. Shah, L. Wallace, and D. Gur, "Detection
and classification performance levels of mammographic masses under different computer-aided
detection cueing environments," Acad. Radiol. 11, 398–406 (2004).
10. R. F. Brem, J. A. Rapelyea, G. Sisman, J. W. Hoffmeister, and M. P. Desimio, "Evaluation of
breast cancer with a computer-aided detection system by mammographic appearance and
histopathology," Cancer 104(5), 931–935 (2005).
11. R. L. Birdwell, P. Bandodkar, and D. M. Ikeda, "Computer-aided detection with screening
mammography in a university hospital setting," Radiology 236(2), 451–457 (2005).
12. A. Lauria, R. Palmiero, G. Forni, M. E. Fantacci, M. Imbriaco, A. Sodano, and P. L. Indovina,
"A study of two different CAD systems for mammography as an aid to radiological diagnosis in the
search of microcalcification clusters," Eur. J. Radiol. 55(2), 264–269 (2005).
13. P. Taylor and R. M. Given-Wilson, "Evaluation of computer-aided detection (CAD) devices,"
Br. J. Radiol. 78, S26–S30 (2005).
14. C. Marx, A. Malich, M. Facius, U. Grebenstein, D. Sauner, S. O. Pfleiderer, and W. A. Kaiser,
"Are unnecessary follow-up procedures induced by computer-aided diagnosis (CAD) in mammography?
Comparison of mammographic diagnosis with and without use of CAD," Eur. J. Radiol. 51(1), 66–72
(2004).
15. P. Taylor, J. Champness, R. Given-Wilson, K. Johnston, and H. Potts, "Impact of computer-
aided detection prompts on the sensitivity and specificity of screening mammography," Health Technol.
Assess 9(6):iii, 1–58 (2005).
16. R. E. Hendrick, G. R. Cutter, E. A. Berns, C. Nakano, J. Egger, P. A. Carney, L. Abraham, S. H.
Taplin, C. J. D'Orsi, W. Barlow, and J. G. Elmore, "Community-based mammography practice: services,
charges, and interpretation methods," Am. J. Roentgenol. 184(2), 433–438 (2005).
258
6.4. Film mammography for breast cancer screening in younger

women is no longer appropriate because of the demonstrated
superiority of digital mammography for this age group
Martin J. Yaffe and Gary T. Barnes

OVERVIEW
The results of a head-to-head comparison of digital vs. conventional film mammographic

screening have been published recently.1 This was a two-year clinical trial conducted by the
American College of Radiology Imaging Network in which about 50,000 women underwent both
digital and film procedures sequentially. The authors of the report claim " that digital
mammography was significantly better than conventional film mammography at detecting breast
cancer in younger women." These findings indicate that digital mammography is justified for
younger patients. Previous studies, however, failed to demonstrate the superiority of digital over
film for mammography. These studies, together with the greatly increased cost of digital over
film, have led some to suggest that it is way too early to abandon film as the modality of choice
for screening young women. Whether or not film is still appropriate for screening younger
patients is the topic of this month's Point/Counterpoint.
Arguing for the Proposition is Martin J. Yaffe, Ph.D. Dr. Yaffe was born in Winnipeg, Canada,
received his B.Sc. and M.Sc. degrees in Physics from the University of Manitoba, and, in 1978,
his Ph.D. in Medical Biophysics from the University of Toronto. He is a Professor in the
Departments of Radiology and Medical Biophysics at the University of Toronto, and serves as
Chair of the Mammography Subcommittee of the AAPM Imaging Committee. His research is
focused on creating improved methods for breast cancer detection. Since 1985 much of his work
has been directed toward the development and validation of digital mammography, as well as on
applications to enhance its value in providing new information to facilitate detection, diagnosis,
treatment, and prevention of breast cancer. He and his wife, Robin, a psychologist, live in
Toronto, where her skills help preserve his sanity.
Arguing against the Proposition is Gary T. Barnes, Ph.D. Dr. Barnes received his B.S. in Physics
in 1964 from Case Institute of Technology (now Case-Western Reserve University), Cleveland,
Ohio and his Ph.D. in Physics in 1970 from Wayne State University in Detroit. Following
completion of his Ph.D., he received postdoctoral training in Medical Physics at the University of
Wisconsin, Madison. In 1972 he joined the Department of Radiology at the University of
Alabama at Birmingham (UAB) where, from 1976 to 1987, he was Chief of the Department's
Physics Section and, from 1987 to 2002, was Director of the Physics and Engineering Division.
In 2002 he became Professor Emeritus. He continues to be involved at UAB part time by
chairing the Radioisotope and Radiation Safety Committee and teaching radiology residents. He
is also involved with X-Ray Imaging Innovations, Inc., a technology development company he
founded in 1998. Dr. Barnes has been active on committees of the AAPM, Southeastern Chapter
of the AAPM (SEAAPM), ACR, RSNA, and the ABR, and is past President of the AAPM (1988)
and SEAAPM (1979). He is a diplomate of the ABR (Radiological Physics), and a Fellow of the
259
AAPM, ACR, and the American Institute of Biomedical Engineers. He was the 2005 recipient of
the Coolidge Award. Dr. Barnes' research interests are in diagnostic x-ray imaging and
mammography and include work on scatter control, screen-film systems, digital radiography, and
clinical medical physics. He is the author or co-author of 10 patents and 100+ scientific papers.
FOR THE PROPOSITION: Martin J. Yaffe, Ph.D.
Opening Statement
When I began working with mammography equipment, it was common for images to be recorded
on direct-exposure film. Presumably this was because of the extremely high spatial resolution
requirements for detecting breast cancer. Doses typically exceeded 10 mGy, timers were
mechanical, proper breast compression was a thing of the future, and automatic exposure control
and grids for mammography did not exist. We've come a long way—through xeroradiography
and dedicated screen-film systems for mammography—and each technical development has
improved the ability to detect breast cancer. Only five years ago, following considerable
collaborative development by several academic and industrial groups, digital mammography was
introduced commercially. The Digital Mammographic Imaging Screening Trial (DMIST)1 may
have been the first time that a new technology for breast imaging was actually put to the acid test
to answer the question: "Is the accuracy of digital mammography equal to or higher than that of
film?" The conclusion was positive—at least for certain subgroups of women, including those
with dense breasts, women under 50, and premenopausal women.
Is digital mammography a mature technology? Of course not—there is ample room for further
improvement and refinement. There is room for growth of new applications that will have a
much greater impact on accuracy than the existence of digital mammography itself-applications
like CAD, tomosynthesis, contrast imaging, telemammography, and dual-energy imaging. Each
of these applications has the potential to improve the detection, diagnosis, and treatment of breast
cancer in different ways. And the use of PACS in conjunction with digital mammography will
improve the efficiency of image management.
So, until we develop something better, digital mammography should become the norm for breast
cancer screening. Women with dense breasts will likely benefit from this technology because of
its wide dynamic range, improved DQE, and adjustable brightness and contrast through a user-
defined lookup table, as well as through sophisticated image enhancement methods that can be
readily applied to digital images.
Another thing we learned from DMIST and other studies is that there is enormous inter-reader
variability in interpreting mammograms. It is clear that technology is only part of the solution.
Just as we have the technology to solve problems like hunger, other factors (like politics and
inertia) frequently block our progress. We will gain at least as much in performance of breast
imaging by improving and standardizing reader skills as we will in converting from film to
digital mammography. What a great idea to do both!
AGAINST THE PROPOSITION: Gary T. Barnes, Ph.D.
Opening Statement
260
The premise for this Point/Counterpoint debate is based on a paper published in the October,
2005 issue of the New England Journal of Medicine.1 A cursory reading of the paper supports the
premise. However, the statement (attributed by Mark Twain to Benjamin Disraeli), "There are
lies, damn lies and statistics," is particularly relevant when one reads the paper more carefully.
Of interest are the sensitivity and specificity values given in Table 3 of the paper and listed in
Table I below.
Inescapable in the paper is the authors' bias for digital mammography. They state: "We found
that digital mammography was significantly better than conventional film mammography for
detecting breast cancer in young women, premenopausal and perimenopausal women, and
women with dense breasts. There was no significant difference in diagnostic accuracy between
digital and film mammography in the population as a whole or in other predefined subgroups." I
find it disturbing that the New England Journal of Medicine reviewers and editor allowed the
authors to make such blanket and unqualified statements. If there is no difference between digital
and film mammography for all women and there is a significant difference in favor of digital
mammography for women less than 50 years old, one might conclude that film mammography
must be better for women greater than 50 years old. Although the greater-than-50-years-old
group was studied, sensitivity and specificity numbers, or for that matter receiver operator
characteristic (ROC) curves, were not given.
Even though sensitivity and specificity values are presented, the majority of the authors'
conclusions are based on comparing areas under ROC curves. Comparing ROC curve areas gives
excessive weighting to regions that have high false positive rates that are not relevant for
screening. For example, the authors conclude that digital is significantly better than film
mammography for dense breasts based on ROC area comparisons. This conclusion is not
supported by the sensitivity and specificity values presented in the paper.
Another concern is that the paper compares the performance of digital mammography units with
existing, and in many instances old and poor performing, film mammography units. The digital
units were all new and tweaked to perform at the highest level. The authors looked for
differences in the units of the four digital-mammography manufacturers that were included in the
study (none were seen), but lumped all film mammography units together. As a reader of
accreditation phantom images for the American College of Radiology (ACR) Mammography
Accreditation Program (MAP) I, and other reviewers have observed that there are very noticeable
differences in image quality of sites that have been accredited by the ACR. There are some
mammography units with better scatter control that consistently yield better accreditation
phantom images than those obtained with other units. The quality of screen-film processing
impacts accreditation phantom and clinical images. The results of the study may well have been
different if only state-of-the-art film mammography units with good scatter control and good
screen-film processing had been compared with digital units.
In summary, the premise that film mammography is inferior to digital mammography and should
not be used in examining young women is built on sand, not stone. The authors of the paper on
which the premise is based are biased. They have chosen the statistics that support their
conclusions, and they do not adequately discuss other statistics that are less supportive of their
bias. It is not clear that the claimed improved diagnostic performance with digital would be
realized if a site has a film mammography unit with good scatter control and good screen-film
processing.
261
Rebuttal: Martin J. Yaffe, Ph.D.
I am disappointed that Dr. Barnes has used one of the classic logical fallacies, the "ad hominem"
argument2—i.e., when you aren't able to attack a position held by a person, attack the person. He
suggests that the DMIST researchers were biased. The DMIST results were published under peer
review in a highly reputable journal. Is he also suggesting that the reviewers were biased? What
is the evidence? In fact, a respected, independent statistical group blinded the researchers to the
data until completion of the analysis. Until the investigators saw these results, they expressed
concerns that a difference between the modalities would not be detectible because of the large
reader variability. While there was hope that digital mammography might prove to be
advantageous, there was a healthy level of skepticism as well. Furthermore, while the initial
policy was to attempt to match doses between the two modalities, the investigators allowed the
doses from digital to drop during the trial based on evidence from physics testing. If anything,
this would work against an attempt to make digital mammography look better than film.
Dr. Barnes also suggests that the digital systems operated in peak form while the film systems
were suboptimal. Again he provides no evidence for this claim. In fact, study sites were selected
for their commitment to high quality and all imaging systems were MQSA compliant. The
performance of both digital and film mammography was carefully monitored. In the case of sites
using photostimulable phosphor systems, film and digital mammography was performed on the
same units. Did the local medical physicists and technologists at all 33 sites work to sabotage
film in comparison with digital? Film mammography is a mature technique, while the steep
learning curve for both manufacturers and radiologists in understanding how best to use the new
technology probably was a factor that could have undermined the success of digital
mammography in DMIST.
Probably of greatest concern, however, is Dr. Barnes' inference that because there was no
statistically significant benefit of digital mammography across the entire study population, while
there was proven benefit for younger women and those with dense breasts, implies that film must
have performed better in older women or those with fatty breasts. This is simply not the case. If
one thinks about the nature of a statistical test, it involves comparing the difference between two
values (in this case, areas under an ROC curve) to the statistical uncertainty in that difference.
The analogy in imaging theory is Rose's criterion for object detectability of a signal difference in
the presence of noise.3 In DMIST, it was found that the difference was significant for younger
women and those with denser breasts. While the difference was also positive (i.e., consistent with
digital mammography being superior) in the overall population, the noise in this difference
caused the results to be insignificant. There was no magic and no deception. Furthermore, the
investigators did not "cherry pick" the data to obtain a result that was significant. The analyses
were preplanned before the study was started. Limitations on article length precluded complete
publication of all analyses, but a full subset analysis will be published shortly.
Finally, Dr. Barnes complains that the areas under the ROC curves were computed using the
entire curve (i.e., using the standard approach). He suggests that this approach weighs the results
inappropriately towards regions of the curve representing low specificity. Although it is
conceivable that some other method might be more relevant, I'm not aware of a better method
that has been validated. I am encouraged that the ROC curves show clear separation in favor of
digital even at low false positive fractions typical of the use of mammography in screening, and
that the curves show no indication of crossing at any specificity level.
262
Digital mammography was developed primarily to address limitations of film for imaging
women with dense breasts, frequently younger women. In DMIST, of the 165 cancers in women
with dense breasts, 24% were found only on digital compared with 11.5% found only on film.1
We should be delighted that mammography can now be more helpful for those women.
Rebuttal: Gary T. Barnes, Ph.D.
I agree with my esteemed colleague that mammography has improved markedly in the past
40 years. I also agree that digital mammography is not a mature technology. I disagree that
digital should be the norm for screening. As noted in my Opening Statement, Reference 1 is
biased. It is not clear that digital is superior to state-of-the-art film mammography. Furthermore,
digital is more expensive. The cost of a digital unit and review workstation is four or five times
that of a conventional film unit and processor. The greater patient throughput of digital units is
compromised by downtime at the sites, at least those I support. Service contracts are expensive
and more than offset the cost of service and of film and chemistry consumed in film
mammography. One inefficiency of digital is that it takes longer for radiologists to read an exam
compared with film. We have a shortage of radiologists reading mammography, and digital
aggravates the problem. For these reasons I suspect that a cost-benefit analysis (based on good
data and science) would show that state-of-the-art film mammography is superior to digital
mammography.
A travesty is that Centers for Medicare & Medicaid Services (CMS) reimburses more for digital
than for film mammography. Reimbursing more for the same study and diagnostic performance
just because the equipment costs more is an absurdity and is caused by lobbying by
manufacturers. Digital mammography is a factor contributing to the spiraling costs of health care
with no benefit.
It is disturbing how poor mammography is, either digital or film, at detecting cancer. A
sensitivity of <50% is not good. We can do better. Based on our experience with CT and the
ability of 3D x-ray techniques to remove superimposed overlying and underlying structures and
improve lesion conspicuity, it is my opinion that 3D x-ray tomosynthesis or cone beam CT
techniques are the future of breast imaging. These will be available in 3–5 years. In addition to
better sensitivity, it is also likely that 3D x-ray imaging will have better CAD performance
(computers as well as humans do better with simpler images).
In conclusion, it is not prudent to invest in a digital mammography system today. If one does,
within a few years it will be necessary to spend a comparable sum once again to purchase a 3D
breast x-ray unit. Be wary of salesmen who promise that a digital unit can be upgraded without
documenting the cost. Two-dimensional projection mammography, whether digital or film, is
nearing the end of its tenure. Wait a couple of years and buy a 3D x-ray unit. It will be a better
investment and worth the wait.
REFERENCES
1. E. D. Pisano et al., "Diagnostic performance of digital versus film mammography for breast
cancer screening," N. Engl. J. Med. 353, 1773–1783 (2005).
2. I. M. Copi, Introduction to Logic (Macmillan, New York, 1994), ISBN 0-02-325041-0.
3. A. Rose, Vision: Human and Electronic (Plenum, New York, 1973), ISBN 0-306-30732-4.
263
TABLE
Table I. Sensitivity and specificity values taken from Table 3 in Ref. 1.
Digital mammography Film mammography
Sensitivity 0.41±0.03 0.41±0.03

All women
Specificity 0.98±0.001 0.98±0.001
Sensitivity 0.49±0.06 0.35±0.06

Women less than 50 years old
Specificity 0.97±0.001 0.98±0.001
Women with heterogeneously Sensitivity 0.38±0.04 0.36±0.04

dense or extremely dense breasts Specificity 0.97±0.001 0.98±0.001
264
CHAPTER 7
Nuclear Medicine
7.1. Positron imaging with SPECT and dual-head scintillation

cameras obviates the need for PET in oncologic imaging
Mark T. Madsen and Beth A. Harkness

Reproduced from Medical Physics, Vol. 25, No. 11, pp. 2116-2118, November, 1998
OVERVIEW
One of the principal clinical applications of positron emission tomographic (PET) imaging of
18F-fluorodeoxyglucose (FDG) has been to evaluate malignant tumors throughout the body.
However, the use of PET is restricted by the cost of dedicated PET scanners and the limited
numbers of PET units (65–70) in the U.S. In place of PET imaging, some centers use single
photon emission computed tomographic (SPECT) cameras equipped with 511 keV collimators to
image the annihilation radiation from FDG. In other institutions, dual-head scintillation cameras
are used for coincidence imaging of the annihilation photons for FDG. One or the other, or
possibly both of these techniques, may have the potential of providing a less costly approach to
tumor imaging with FDG, when compared with PET.
Arguing for the Proposition is Mark Madsen. Dr. Madsen is an Associate Professor of Radiology
at the University of Iowa, Iowa City, IA. His research interests include image processing,
internal radiation dosimetry, and quantitative SPECT and PET imaging. Dr. Madsen received his
Ph.D. in Radiological Sciences from the University of Wisconsin in 1979. He served as the chair
of the AAPM Nuclear Medicine Committee from 1986 to 1989, has participated in and chaired
several AAPM task groups, and is the current chair of the AAPM Special Interests Group
Committee. Dr. Madsen also organized the 1997 AAPM/RSNA Physics Tutorial for Residents.
Arguing against the Proposition is Beth Harkness. Beth Harkness is a medical physicist in the
PET Center and an Assistant Professor of Radiologic Sciences at Wake Forest University School
of Medicine in Winston- Salem, NC. Ms. Harkness has a degree in radiation science (M.S.
Georgetown University, 1992) and is board certified by the American Board of Radiology in
Medical Nuclear Physics. Ms. Harkness has been working in the field of nuclear medicine since
1977. During the first 15 years of that time her primary area of interest was SPECT imaging, but
she has been working in PET since she joined the faculty at Wake Forest in 1992.
265
For the proposition: Mark T. Madsen
Opening Statement
This debate concerns the current and future role of positron emission tomography (PET)
instrumentation in diagnosing cancer. The two choices currently available are dedicated PET
systems and the SPECT/PET hybrid systems. Dedicated PET systems are specifically designed
for coincidence imaging of 511 keV annihilation photons, while SPECT/PET hybrid systems are
scintillation camera based SPECT systems with the added capability of coincidence imaging. 1
There are many compelling reasons why the hybrid systems will become the primary instruments
used in oncologic imaging:
Efficacy. The current interest in PET imaging is not based on the performance of dedicated PET
systems, but on the efficacy of 18F-FDG. This compound is the premier radiopharmaceutical for
tumor imaging in nuclear medicine today, with demonstrated high sensitivity and specificity for
diagnosing lung, breast, head and neck, pancreatic and colon cancer as well as melanoma and
lymphoma.2 Because 18F-FDG concentrates strongly in these tumors, the performance
requirements of imaging systems are not quite as critical as they would be for other
radiopharmaceuticals. SPECT imaging of 18F-FDG with high energy collimators has been shown
to be effective for lesions >2 cm. 3–5 In a recent study, the hybrid systems operating in the
coincidence mode found 93% of the lung lesions and 65%–70% of the mediastinal and neck
lesions detected by dedicated PET systems. 6 This performance will likely be improved as hybrid
systems evolve with more efficient detectors and better reconstruction algorithms.
Flexibility. Dedicated PET systems can only produce images of positron emitters and will be
idle otherwise. The SPECT/PET hybrid systems can be operated as full performance SPECT
systems with collimators, or the collimators can be removed for coincidence imaging. For
oncologic imaging, there are advantages to both approaches. Because the count sensitivity and
spatial resolution associated with high energy collimators are poor, one normally would not
resort to collimators if a coincidence option were available. However, collimated systems permit
the simultaneous acquisition of multiple energy windows so that 18F-FDG can be imaged along
with other radiotracers such as 99mTc-HDP, a bone scan agent. In addition, the hybrid system
could be used as a conventional SPECT system during times when PET studies are not requested
or PET tracers are unavailable. The thicker detectors used in hybrid systems provide the added
advantage of improved count sensitivity for imaging 67Ga, 111In, or 131I.
Cost. It is clear that one can not make an honest case for the hybrid system from strict image
performance arguments. Although hybrid systems operating in the coincidence mode have
comparable spatial resolution, dedicated PET systems have much better detection efficiency and
count rate performance. Thus, for comparable imaging times dedicated PET systems should
allow better detection of small, low contrast lesions. This is borne out by the results of
comparison studies that show that hybrid systems miss about half or more of the lesions <1 cm
that are seen on dedicated PET systems.
However, the cost of a dedicated PET system exceeds one million dollars, putting its purchase
out of reach of many institutions. 7 PET applications in oncology will not thrive until many
institutions have access to both PET radiopharmaceuticals and PET imaging systems. Currently
there are only 65–70 medical institutions in the U.S. with a dedicated PET system. This is
obviously too small a number to serve the healthcare needs of the patient population. It is also
266
too small to provide referring physicians with direct experience in PET imaging. With the
current economic constraints imposed on health care organizations, it is unlikely that enough
dedicated PET systems will be sold in the near future to rectify the situation. The cost is simply
too high and the risk too large (and also real; several PET centers have stopped operation over
the last 5 years). Hybrid systems by comparison are attractive since they deliver 70%–80% of the
performance of a dedicated PET system at less than half the cost.
Rebuttal
The superior performance of dedicated PET systems was acknowledged in my opening

statement. If that were the only issue, there would be no debate. However, the growth of PET has
been painfully slow because of the cost and inflexibility of dedicated positron systems. The
success of PET imaging in clinical oncology requires the access provided by the hybrid systems.
REFERENCES
1. P. H. Jarritt and P. D. Acton, ‗‗PET imaging using gamma camera systems: a review,‘‘
Nucl. Med. Commun. 17, 758–66 (1996).
2. C. K. Hoh et al., ‗‗PET in oncology: will it replace the other modalities?,‘‘ Seminars
Nucl. Med. 27, 94–106 (1997).
3. D. J. Mcfarlane et al., ‗‗Triple-head SPECT with 2-@fluorine-18#fluoro- 2-deoxy-D-
glucose (FDG): initial evaluation in oncology and comparison with FDG PET,‘‘ Radiology 194,
425–429 (1995).
4. W. H. Martin et al., ‗‗FDG-SPECT: correlation with FDG-PET,‘‘ J. Nucl. Med. 36, 988–
995 (1995).
5. W. H. Martin et al., ‗‗Detection of malignancies with SPECT versus PET, with 2-
@fluorine-18#fluoro-2-deoxy-D-glucose,‘‘ Radiology 198, 225–231 (1996).
6. P. Shreve et al., ‗‗Onocologic diagnosis with 2-@fluorine-18#fluoro-2- deoxy-D-glucose
imaging: dual-head coincidence gamma camera versus positron emission tomographic scanner,‘‘
Radiology 207, 431–437 (1998).
7. P. S. Conti, J. S. Keppler, and J. M. Halls, ‗‗Positron emission tomography: a financial
and operational analysis,‘‘ Am. J. Roentgenol. 162, 1279–1286 (1994).
Against the proposition: Beth A. Harkness
Opening Statement
The imaging task must be considered when determining if SPECT with a high energy collimator
(HE-SPECT) or camera-based coincidence imaging (Camera-CI) are adequate replacements for
positron emission tomography with dedicated scanners (True-PET). In oncologic imaging we are
trying to evaluate a known mass for the presence of malignancy (primary or recurrent) and,
often, the stage of the disease by identifying new lesions. True-PET has been shown to perform
these tasks more reliably than other modalities (CT and MRI) for a number of cancers (e.g., lung,
colon, breast, and brain). Do HE-SPECT or Camera-CI perform as well for these clinical tasks?
Determining if a known mass represents a malignancy requires high specificity. Specificity is
governed by the amount of noise in the image. If an image is noisy, the differentiation between
normal and abnormal tissue is more difficult. HE-SPECT and Camera-CI (even with a 5/8 in.
NaI(Tl) crystal) have much lower quantum efficiency than True-PET. Additional factors such as
the collimator and count-rate limitations of the gamma camera further limit the capability of HE-
267
SPECT and Camera-CI. The higher efficiency of True-PET produces images that contain lower
noise levels and thus higher specificity.
Identification of a malignancy where it is not already identified by another modality, as in the

case of metastatic spread to lymph nodes, requires a test that has high sensitivity for small
lesions. This is directly dependent on the resolution and contrast. True-PET is equal to or
superior to HESPECT and Camera-CI for each of these parameters. The system resolution of
HE-SPECT, Camera-CI, and True-PET are about 20, 4.5, and 4.5–6 mm, respectively. The
spatial resolution HE-SPECT is inferior to True-PET, and Camera- CI, and investigators have
shown that HE-SPECT is not useful for evaluating tumors less than 2 cm in diameter. The spatial
resolution of True-PET and Camera-CI are about equal and one may expect the contrast to be
similar. Camera-CI uses a three-dimensional (3D) acquisition which increases system sensitivity
but results in the acquisition of a large number of scatter and random events. If uncorrected or
partially corrected, the scatter and random events appear as background counts, thus reducing
contrast. Some manufacturers are now using shields on Camera-CI systems to decrease
the amount of scattered radiation detected. This further decreases the number of coincidence
events detected and increases image noise.
Where it is important to be able to determine if a mass is benign or malignant, it is at least as

important to accurately stage the disease. The ability to identify small foci of increased uptake is
a major factor in the success of True-PET. HE-SPECT and Camera-CI have inherent limitations
that result in low sensitivity for these small lesions ( 1.5 cm)1 which is a significant limitation in
oncologic imaging. Identification of malignant lymph nodes called normal or indeterminate by
CT criteria is where True-PET has excelled and established a role in oncologic imaging.
Rebuttal
The efficacy of 18F-FDG imaging in oncology is based completely on the use of True-PET
systems. The sensitivity and specificity are dependent on the ability to identify and characterize
areas of increased 18F-FDG uptake. Low resolution and increased noise that are a function of the
detection device will decrease the ability to perform these tasks. A truly useful diagnostic test
should have both high sensitivity and specificity.
The desire to perform PET imaging with HE-SPECT or Camera-CI at a lower initial cash
investment can be tempting but may be unwise in the long run. While these methods are
adequate for some limited applications (i.e., detection of large tumors) they are severely
compromised in others (staging and detection of small tumors). A particular example is the use
of Camera-CI in detection of mediastinal and neck lesions mentioned by Dr. Madsen. He points
out that Camera-CI detected 65%–70% of the lesions detected by PET. This might be considered
adequate if PET had 100% sensitivity, but in reality the sensitivity of True-PET for lesions in the
mediastinum is around 80%. This makes the sensitivity of Camera-CI about 50%–55%. Where
the specificity using Camera-CI for these lesions may be very high, we would miss nearly 50%
of the metastatic lymph nodes. How long would it take referring physicians to stop using a test
that failed to detect significant disease 50% of the time?
The 1–2 million dollars price for a PET scanner is a lot for nuclear medicine equipment, but in
reality is quite comparable to the price of CT and MRI scanners. This seems to be a reasonable
price for improved diagnostic information for oncology patients.
268
REFERENCE
1. P. D. Shreve, R. S. Steventon, E. C. Deters, P. V. Kison, M. D. Gross, and R. L. Wahl,

‗‗Oncologic diagnosis with 2-@Fluorine-18#Fluoro-2-deoxy-D-glucose imaging: Dual-head
coincidence gamma camera versus positron emission tomographic scanner,‘‘ Radiology 207,
431–437 (1998).
269
7.2. Isotopic labeling with short-lived radionuclides represents the

future for clinical PET imaging
Beth A. Harkness and Mark T. Madsen

OVERVIEW
PET imaging in radiation oncology is performed today exclusively with 18F-deoxyglucose

(FDG). The 18F half-life is long enough to permit its delivery from a regional supplier, so that an
on-site cyclotron is not required. Many nuclear imagers see FDG as the radiopharmaceutical of
choice for oncologic imaging well into the future. Others believe that isotopic labeling with
biologically-active molecules tagged with short-lived 11C, 13N and 15O as the real future of PET
imaging in oncology. This difference of opinion is explored in this month's Point/Counterpoint.
Arguing for the Proposition is Beth Harkness, M.S. Ms. Harkness is the nuclear medicine
physicist for Henry Ford Health System in Detroit, MI. She received a master's degree in
radiation science from Georgetown University in 1992, and is board certified by the American
Board of Radiology in Medical Nuclear Physics. Her primary areas of interest are PET and
SPECT imaging. She was involved in the early development of SPECT as a clinical tool, and has
also spent 10 years working as a physicist in a PET center. She has been a member of the Nuclear
Medicine (1994–2000) and the Remotely Directed Continuing Education (1998–2002)
Committees of the AAPM.
Arguing against the Proposition is Mark T. Madsen, Ph.D. Dr. Madsen is an associate professor
in the Department of Radiology at the University of Iowa. He received his Ph.D. degree in 1979
from the University of Wisconsin for investigating positron emitting regional cerebral blood flow
agents. He was on the faculty at Thomas Jefferson University in Philadelphia from 1979 until
1988 when he moved to the University of Iowa. He is a current member and a past chair of the
AAPM nuclear medicine committee and the current chair of the ABR medical nuclear physics
subcommittee. Dr. Madsen's research interests include image reconstruction techniques and
quantitative SPECT and PET imaging.
FOR THE PROPOSITION: Beth Harkness, M.S.
Opening Statement
The current trend in nuclear medicine is for departments to purchase PET scanners and use [18F]-
2-fluoro2-deoxy-D-glucose (FDG) purchased from a nuclear pharmacy to image patients.
Oncologic, neurologic, and cardiac diseases are currently being imaged using this
radiopharmaceutical. The boom in PET oncology imaging is due, in part, to the fact that FDG is
readily available. For nuclear medicine to emerge as a "metabolic imaging" discipline, however,
radiotracers will be needed that are target specific and that follow the physiologic pathway being
evaluated. FDG does not meet either of these objectives. In the vascular space, FDG crosses the
cell wall via facilitated transport using the glucose transport proteins in a manner analogous to
270
glucose. In the cell, FDG is converted to FDG-6-phosphate in the same manner that glucose is
converted to glucose-6-phosphate. But this is where the similarity ends. Glucose-6-phosphaste
proceeds along the glycolytic pathway, but the cell cannot process FDG-6-phosphate. In addition,
the uptake pathway is the same for all cells, whether they are normal or cancerous, although there
is more radiotracer uptake in the latter. One would describe FDG as a nonspecific radiotracer.
To become the metabolic imaging modality, nuclear medicine will have to use radiotracers that
are specific in their uptake mechanisms and that are metabolized in the same manner as their
nonradioactive counterparts. The need to use short-lived radionuclides, such as 11C is obvious.
Because carbon is found in many biologic molecules, 11C can be incorporated isotopically into
these molecules without changing their biologic function. There are numerous examples of this
approach, including 11C-acetate for imaging fatty acid metabolism in the heart, and 11C-choline
and 11C-labeled amino acids for tumor imaging. Even 11C-glucose would be better than FDG for
evaluating metabolic function in different cell types. In tumors, it may be advantageous to
classify or stage the aggressive nature of a tumor by evaluating the uptake and clearance
characteristics of 11C-glucose, compared with what we are doing now with FDG.
The need to produce short-lived tracers will, in the future, motivate manufacturers of cyclotrons
to produce turn-key systems that are easy to operate. History tells us that what is difficult or
nearly impossible to do today will become commonplace tomorrow through the work of
innovative scientists. The fact that it is difficult to obtain 11C tracers outside the research
environment today does not mean that this will remain a problem in the future. As investigators
show that short-lived metabolic analogues are indeed better for the evaluation and diagnosis of
various diseases, cyclotrons in the nuclear medicine department will become common.
Rebuttal
No one would dispute the statement that the cost of producing short-lived radionuclides such as
11
C, 15O, and 13N is high, and that the process of approval for marketing and cost-reimbursement
has been long. This statement needs to be examined in the light of trends in medical imaging.
The first issue is cost. It is true that the cost of a cyclotron is about $2,000,000, and that, in
addition, personnel are required to operate the cyclotron and produce radiopharmaceuticals. One
has to examine this cost relative to other costs in medical imaging. Most high-end imaging
equipment costs 2–3 million dollars. A good example is a PET/CT scanner. The nuclear medicine
community perceives that combining a CT scanner with PET adds diagnostic value to the PET
oncology scan. Thus, nuclear medicine physicians and oncologists are willing to commit the
funds needed to purchase a combined unit, and to hire the personnel, at least two technologists, to
operate the equipment. Many PET/CT scanners are being purchased even though the
reimbursement for the CT portion of the study is unresolved. Why? Because the physician
perceives that there is a clear diagnostic advantage to using the CT scan for attenuation
correction, and for interpreting the spatially co-registered PET/CT study. Cost is not an issue if
the technology is perceived to improve diagnostic accuracy and patient care.
The second issue is the difficulty in obtaining FDA approval of new radiopharmaceuticals.
Approval of new radiopharmaceuticals has been a troubled process for the nuclear medicine
community for many years. With the FDA Modernization Act of 1997, Congress attempted to
address the approval time of pharmaceuticals, including radiopharmaceuticals, that has plagued
nuclear medicine for the past 25 years. This act directly addresses problems of PET
pharmaceutical approval by instructing the FDA to have different guidelines for not-for-profit
271
medical institutions that produce them. 18FDG is an example of both the difficulty and ultimately
successful approval. It took many years to obtain the first approved application of 18FDG, but
now there are several approved applications in oncology. Not every facility now must submit a
NDA or INDA. The growth we are currently experiencing in PET is through the combined efforts
of many professional organizations to obtain approval for 18FDG.
There is no reason this could not occur for short-lived radiopharmaceuticals. Finally, it is the
history of medicine that academic centers lead the way in development of new technology. The
future of PET lies in the continued development of PET tracers, including those using short-lived
radionuclides.
AGAINST THE PROPOSITION: Mark T. Madsen, Ph.D.
Opening Statement
This is a debate where the economic realities of medical imaging play a decisive role. No one can
reasonably argue that there is any disadvantage of exploiting the wide potential of
radiopharmaceuticals offered by 15O, 13N and 11C, except that the cost will be unrealistically high
with little chance of ever recovering a significant fraction of the expense. Short-lived
radionuclides will require onsite production and compounding, and the capital equipment,
personnel and regulatory expenses associated with these activities will be very high.1 Let's
examine these expenses.
Capital equipment. Radionuclides with half lives under 1 hour require an onsite cyclotron. Onsite
means not just somewhere within the institution, but in close proximity to the imaging facility. In
addition to the large capital equipment cost for the cyclotron and the site preparation (on the
order of $2,000,000), the recurring costs of cyclotron maintenance and the associated utilities are
substantial. Onsite compounding of PET radiopharmaceuticals requires space for, and the
installation of, specialized equipment such as hot cells, fume hoods, dose calibrators, gas and
high performance liquid chromatographs, shielding, and survey meters. The cost for this
equipment can be well over $250,000.1
Personnel. Highly trained individuals are required to operate the cyclotron and compound the
radiopharmaceuticals. Although cyclotrons are largely computer controlled, they still require a
portion of an individual's time to operate them. In addition, there are substantial radiation safety
issues associated with a cyclotron that are time consuming. Compounding of the
radiopharmaceuticals is even more time consuming, because it includes not only the daily
preparation of the radiotracers, but all of the associated quality assurance and paperwork.
Combined together these responsibilities significantly exceed 1 FTE.1
Regulatory requirements. Radiopharmaceuticals that are used clinically in humans require either
the approval of a New Drug Application (NDA) or an Abbreviated New Drug Application
(ANDA) from the Food and Drug Administration (FDA). 2 Preparation of the paperwork for
either an NDA or ANDA is extremely time consuming and could easily account for 0.5 FTE of a
radiochemist or radiopharmacist. The application fees are substantial (>$200,000 for an NDA)
and there are yearly recurring fees that exceed $10,000. Combined with the radiation safety
demands, the regulatory burden can easily exceed $50,000/year.1
272
The combined cost of the required capital equipment, personnel and regulatory compliance make
it extremely difficult to recover expenses unless there is substantial reimbursement. Obtaining
reimbursement for any radiopharmaceutical is often a long and difficult struggle, even when it is
approved by the FDA. It took years of intense lobbying to get 18F-FDG reimbursed, in spite of
the large body of evidence supporting its efficacy. The April 1991 issue of the Journal of Nuclear
Medicine was devoted to clinical PET and the role of 18F-FDG imaging,3 yet Medicare
reimbursement for the first 18F-FDG applications did not occur until 1998.4 There is no reason to
assume that the process will be any easier for other PET radiopharmaceuticals, especially when
their initial use will be limited to academic centers that have the equipment and personnel to
support the effort.
Rebuttal
I have no disagreement with Ms. Harkness about the potential of short-lived PET radiotracers to
provide unique diagnostic information. The likelihood is extremely low, however, that healthcare
providers will invest in the equipment and personnel necessary for on-site manufacturing of PET
radiopharmaceuticals. Innovations may reduce equipment costs for production of short-lived
radionuclides, but are unlikely to have much effect on the recurring costs of personnel and
regulatory compliance. Ms. Harkness suggests that useful information may be gleaned from
examining the kinetics of short-lived PET radiotracers. Although this is undoubtedly true, the
acquisition and processing of these studies require substantial equipment and personnel time.
This requirement adds to the already prohibitive expense of PET studies with short-lived
radionuclides.
For a facility to overcome the economic burdens of using short-lived PET radionuclides, the
following conditions must be met: (1) The labeled compound must have a wide range of
applications, as does 18F FDG. Radiotracers that cannot be used on a large patient population
cannot recover the development and operational costs and are also unlikely to be reimbursed. (2)
If a new radiotracer is developed that does have wide applicability, it will have to be impossible
to label it with 18F. Otherwise, market forces will selectively remove the short-lived alternative.
For example, interest in 11C choline as a potential clinical agent was significantly dampened
when 18F fluorocholine became a reality.
With all these considerations, it is reasonable to conclude that short-lived PET radionuclides are
not a viable option for most healthcare providers, neither now, nor in the future.
REFERENCES
1. J. S. Keppler and P. S. Conti, "A cost analysis of positron emission tomography," AJR,
Am. J. Roentgenol. 177, 31–40 (2001). M. Rotman, D. Laven, and G. Levine,
"Radiopharmaceutical regulation and Food and Drug Administration policy," Semin. Nucl. Med.
26, 96–106 (1996).
2. H. N. Wagner, Jr., "Molecular medicine: From science to service," J. Nucl. Med. 32,
561–564 (1991).
3. W. J. Smith, "Revisions to payment policies under Medicare physician fee schedule
proposed for 1999," J. Nucl. Med. 39, 28N–30N (1998).
273
7.3. Correction for image degrading factors is essential for accurate

quantification of brain function using PET
Habib Zaidi and Vesna Sossi

OVERVIEW
Quantification of brain function using positron emission tomography (PET) is an emerging

research field. For many years, interest in the regional distribution of brain perfusion and
metabolism has steadily increased, primarily because many neurological disorders are associated
with a decrease of regional cerebral blood flow (rCBF). PET offers the possibility of quantitative
measurements of tracer concentration in vivo. However, there are several issues that must be
considered in order to fully realize this potential. In practice, the measured line integrals must be
corrected for a number of background and physical effects, the three most significant being the
limited spatial resolution and associated partial volume effect, photon attenuation, and the
contribution of events arising from photons scattered in the object and gantry. Many investigators
believe that accurate patient-specific correction for these image-degrading factors is essential for
quantification of relevant parameters when imaging brain function using PET. Others often
neglect them or believe that approximate correction methods can be used with confidence. This
difference of opinion is explored in this month's Point/Counterpoint.
Arguing for the Proposition is Habib Zaidi, Ph.D. Dr Zaidi is senior physicist and head of the
PET Instrumentation & Neuroscience Laboratory at Geneva University Hospital. His research
centers on modeling nuclear medical imaging systems using the Monte Carlo method, dosimetry,
image correction, reconstruction and quantification techniques in emission tomography and
functional brain imaging, and more recently on novel design of dedicated high-resolution PET
scanners in collaboration with CERN. He is Associate Editor for Medical Physics, a member of
the editorial board of Computer Methods and Programs in Biomedicine and the International
Journal of Nuclear Medicine, and scientific reviewer for several medical physics, nuclear
medicine and computing journals. He is affiliated to several medical physics and nuclear
medicine organizations and member of the professional relations committee of the IOMP.
Arguing against the proposition is Vesna Sossi, Ph.D. Dr. Sossi is Assistant Professor of Physics
and Astronomy at the University of British Columbia (UBC), and head of the physics program of
the UBC/TRIUMF PET group. She received her Ph.D. in Nuclear Physics from UBC and
changed research fields to PET imaging immediately after her degree. She is actively involved in
PET instrumentation and data quantification and reconstruction research, with particular
emphasis on quantitative 3-dimensional high resolution brain imaging. She was involved in the
early characterization of hybrid PET/SPECT imaging, and was part of the review committee for
the 2001 NEMA PET standards. She is a scientific reviewer for several medical physics journals.
FOR THE PROPOSITION: Habib Zaidi, Ph.D.
Opening Statement
274
During the last decade, neuroimaging has advanced elegantly in the medical and research arenas.
PET, with its superior sensitivity and spatial resolution, appears uniquely suited to take the lead
in this promising field of imaging. Convincing clinical evaluations and research investigations of
PET are providing clinicians and neuroscientists with relevant functional information in various
pathologies including cerebrovascular disorders, brain trauma, epilepsy, dementia, Parkinson's
disease and brain tumors, and in mental disorders such as depression, schizophrenia and
obsessive-compulsive disorders. Within the context of functional brain imaging, the aim of
quantification is to provide a reliable numerical measure of brain function. For quantitative
analysis of PET images, several image-degrading effects must be accounted for, including poor
signal-to-noise ratio, limited spatial resolution, and spatially-varying loss or corruption of signal
due to photon interactions with matter. Photon attenuation and contributions from scattered
photons reduce the accuracy of measured activities and activity concentrations. 1,2 In addition,
limited spatial resolution causes an object to appear enlarged if its true size is less than 2–3 times
the system resolution. While the total reconstructed counts within the object are conserved, the
count density is decreased from the true value because the data are "smeared" over a larger area.
This characteristic is known as the partial volume effect.
Advances in quantification of brain function (blood flow, metabolism and receptor

characteristics) with PET, especially for small structures such as the putamen and hippocampus,
rely on two improvements: (1) hardware improvements to enhance spatial resolution and
sensitivity, and addition of components to correct for degrading factors (e.g., electronically-
delayed coincidence windows for subtraction of random coincidences, and transmission scanning
for patient-specific attenuation correction); and (2) software improvements to attain better image
quality and achieve more accurate quantification of relevant parameters. Significant progress has
been made in the design of high-resolution 3D PET units with the capacity to acquire more
accurate depth-of-interaction information.3 Also, improved reconstruction algorithms have been
proposed, and some have been incorporated into software supplied by manufacturers to end-
users. In spite of improved algorithms for accurate smoothing of random coincidences,
attenuation correction and scatter compensation, and partial-volume effect-correction, however,
many PET users still do not correct for these factors, either because they lack confidence in the
algorithms, or they do not appreciate the importance of the corrections owing to the absence of
physics support in their institutions.
Earlier PET studies inconsistently reported aging-induced reductions in CBF, oxygen

metabolism, and glucose metabolism. Metzler et al.4 have shown that this observation may
reflect the absence of correction for the dilution effect of age-related cerebral volume loss. The
traditional approach to accounting for the partial volume effect in the quantification of brain PET
images uses anatomical information derived from magnetic resonance imaging (MRI) and the
spatial resolution characteristics of the PET unit. Although considered as a limiting factor, MR
images are generally available for patients undergoing cerebral PET scanning through hospitals'
picture archiving and communications systems. Moreover, it has been shown recently that MRI
data can be used for attenuation and scatter correction purposes in 3D brain PET imaging using a
transmission less scanning protocol.2 Reconstruction of PET images without attenuation
correction can cause excessive count densities and reduced image contrast in regions of low
attenuation. Scatter qualitatively decreases contrast by misplacing events during reconstruction,
and quantitatively causes errors in the reconstructed radioactivity concentrations by
overestimating the actual activity. All of these effects can introduce artifacts into radionuclide
images that complicate visual interpretation and cause profound errors in quantification. For
these reasons, it is essential to understand both the physical processes that underlie the data
acquisition process, and the methods that can be used to correct images for inaccuracies. These
275
correction methods are now widely accepted by the nuclear medicine community as essential for
achieving artifact-free, quantitatively accurate data.
In general, there is no rational motivation why sophisticated correction methods for all of the
physical degrading effects should not be applied to brain PET images prior to extraction of
relevant quantitative data in a clinical and research environment.
Rebuttal
I agree with Dr. Sossi that accurate quantification requires extensive technical and organizational
efforts that may be unaffordable for a clinical department with limited scientific support. The first
question to be answered is "What is expected from such studies?" Investments should be
comparable to expectations. The second interesting question is "Would one expect similar results
between images obtained with and without correction for the physical degrading effects?" The
answer would provide a true comparison of the effect of different correction techniques on
relevant quantitative parameters when studying brain function using PET. Such an experiment
would be difficult to perform with clinical data but should be easily performed using either
experimental phantom measurements or Monte Carlo simulation studies, which have the
advantage of being able to generate data sets in a controllable manner and switching on and off
the effect of the physical degrading factors. Most studies concluded that correction methods
improved the quantitative accuracy compared to the case where no corrections were applied. 1 No
one would dispute the statement that significant progress in quantitative PET imaging has been
made over the last few years as a result of improved correction methods for attenuation, scatter
and partial-volume effect. The specific benefits of transmission-based attenuation correction, in
contrast with calculated attenuation correction, are the subject of heated debate. 2 Contribution of
scatter from outside the FOV remains a challenging issue that needs to be addressed carefully in
whole-body imaging especially with large axial FOV 3D PET units. However, this is a less
challenging issue in brain scanning. In PET activation studies characterized by low count
statistics, subtraction-based scatter correction methods add considerable noise, which jeopardizes
the significance of statistical analysis. This problem has been tackled with iterative
reconstruction-based scatter compensation techniques, where the scatter component is modeled
within the projector/backprojector pair. This approach results in better noise properties than
direct subtraction. Accurate scatter modeling has recently been achieved using computationally
efficient fully 3D Monte Carlo simulation-based reconstructions.5 In addition, correction for
partial volume effect might influence the results of statistical analysis in group comparisons.
Using Statistical Parametric Mapping analysis performed on subjects with probable Alzheimer's
disease and age-matched healthy volunteers, Matsuda et al.6 have shown that the significance of
the rCBF decrease in the bilateral amygdala and hippocampi disappeared after correction of
partial volume effect, while the significant decrease in the bilateral parahippocampal gyri
remained.
Reconstruction methods are continuously being improved, and scanner manufacturers are
optimizing the performance of dedicated software by integrating latest algorithmic developments.
Maximum a posteriori reconstructions using a Bayesian model in combination with a Poisson
likelihood function and a Gibbs prior on the image provide images of higher resolution. The
value of improved models to correct for attenuation, scatter, and partial-volume effects,
performed on raw projection data, preliminary reconstructions, or integrated with the transition
matrix of an iterative reconstruction algorithm, is still an open question and remains a good
academic problem in functional brain imaging. This question requires further research and
development.
276
AGAINST THE PROPOSITION: Vesna Sossi, Ph.D.
Opening Statement
From an idealistic point of view, absolute quantification is certainly desirable to obtain a truthful
representation of the biological process or metabolic function being imaged with PET. In the real
world, however, we are faced with a variety of limitations, starting with the observation that PET
radioactive decay is statistical in nature and therefore cannot be precisely determined. In
addition, photons interact with all forms of matter, instrumentation is capable of only limited
spatial, energy and temporal resolution and requires careful calibration, and patient motion
during a scanning procedure produces blurred images. Correction algorithms exist for most of
these effects. There are, however, three questions that must be addressed: What is the overall cost
of applying the corrections, are the available corrections sufficiently accurate to increase the
quantitative accuracy of the results, and, most importantly, are the corrections really necessary
for every study in order to obtain clinically meaningful results? To answer these questions, four
main sources of accuracy loss will be examined: attenuation, scatter, presence of random events
and the partial volume effect.
Photon attenuation has been repeatedly identified as the single most important factor in the loss
of quantification ability. However, attenuation corrections have also been identified as a
significant source of errors because measured attenuation corrections can introduce additional
noise in the images, thus reducing the image contrast, which is especially relevant in tumor
imaging. Also, a mismatch between the spatial location of the subject during the emission and the
transmission scans can introduce significant artifacts into attenuation-corrected images. Finally,
the presence of metallic dental implants can introduce artifacts into brain images, not only when
CT is used to determine the attenuation correction coefficient, as in new PET/CT scans, but also
when a standard positron source is employed for attenuation correction.7
Another aspect to be considered is that addition of a transmission scan generally lengthens the
scanning procedure and increases the dose of radioactivity that must be administered to the
patient. For example, in 18F-flurodeoxyglucose (FDG) PET imaging, a tracer uptake period is
generally required after tracer injection and before initiation of scanning. If a post-injection
transmission scan is not feasible, the patient must undergo a transmission scan first, and then
either lie on the scanning bed for a considerable time before the emission scan is started, or leave
the bed and then be carefully repositioned for the emission scan. In addition to reducing the
useable scanner time, both options increase the risk of a position mismatch between the emission
and transmission scan.
The clinical utility of this correction must also be addressed. Certainly there are PET studies
where attenuation correction is essential such as determination of process rate constants with
methods that use plasma-derived input functions. There are other applications, however, where
attenuation correction is not only unnecessary, but even reported as detrimental. Bengel et al.8
argue, for example, that nonattenuation corrected FDG images yield improved contrast between
tumor and background. For head and neck tumors, for example, the contrast in nonattenuation
corrected images is approximately twice that in attenuation corrected images.
Detection of scatter events is another major contributor to quantification loss. In 3D brain

imaging, the fraction of scatter to total detected events can be as high as 40%. No fully accurate
scatter correction methods are currently available. Available methods are based on scatter
277
modeling using physics principles, and often fail to account for scatter originating from
radioactivity outside of the field of view. The scatter fraction depends only on the thickness of
material traversed by the photons; it does not depend on the injected dose or radiotracer
distribution. Therefore, scatter only minimally affects the comparison of images obtained with
the same radiotracer in the same patient.
The detection of random events adds a fairly uniform background across the field of view, and
thereby reduces image contrast. In contrast to scatter, however, the number of random events is
count-rate dependent. If the objective of the study is to compare between two conditions, the
effect of random events can be minimized by maintaining similar count rates among scans.
The limited spatial resolution of the scanner causes the partial volume effect, which affects the
estimate of radioactivity concentration for all objects that are smaller than the tomographic
resolution element. This is often the case in brain imaging. Several partial-volume correction
algorithms exist, but they involve many processing steps and generally require acquisition of an
MRI scan to define the anatomical size of the structures involved in the function being
investigated with PET. These additional steps are costly in expense and time, and may introduce
further errors. For instance, co-registration between PET and MRI images has an accuracy limit
of approximately 2 mm, which is fairly large considering that structures of only a few mm are
often of interest.
Frequently, PET studies focus on identification of functional differences between subjects

scanned under different conditions. In these cases the partial volume effect is approximately
constant between the two conditions, since inter-subject comparisons are used, and the
distribution of radiotracer may not change greatly. In such studies differences of 10–20% in the
kinetic parameters derived from the study are often found to be significant.9 The magnitude of
the partial-volume correction may cause some parameter values to increase several fold, with an
associated increase in noise.10 Application of such a correction, if not exact, would jeopardize the
ability to detect subtle biological effects. Finally, the effect of image blurring caused by patient
motion on the accuracy of the partial-volume correction remains to be investigated.
Rebuttal
Dr. Zaidi is absolutely correct when stating that better image quantification has dramatically
improved the investigative power of PET and has contributed to more accurate biological
discoveries. He correctly argues that significant advances have been made both in software and
hardware associated with PET scanners that yield far more accurate results. As he points out,
many of the correction methods require sophisticated measurements and software techniques.
This enhanced sophistication requires a high degree of precision and expertise that may not
always be available. If imprecise, the correction methods might introduce additional sources of
artifact and noise. Correction methods can be implemented only if all sources of potential errors
have been thoroughly investigated in the context of the particular scanning protocol. This
includes factors often neglected, such as the effects of patient motion on the accuracy of the
corrections.
Another aspect to consider is the practicality of obtaining particular correction factors. For
example, most of the partial-volume correction methods require MR imaging, which significantly
increases the cost of the scanning procedure and the burden to subjects undergoing a clinical
examination. It is important to question how much additional information will be acquired by
278
implementing the correction methods. The answer is likely to be different for different types of
studies or clinical examinations and for different scanning environments.
In summary, pursuit of absolute quantification should undoubtedly continue, because it provides

a more truthful representation of the processes being imaged. It is also important, however, to
evaluate on an individual case basis if the additional accuracy potentially provided by the
quantification procedures will significantly benefit the outcome of a PET study.
REFERENCES
1. H. Zaidi, "Comparative evaluation of scatter correction techniques in 3D positron

emission tomography," Eur. J. Nucl. Med. 27, 1813–1826 (2000).
2. H. Zaidi and B. H. Hasegawa, "Determination of the attenuation map in emission
tomography," J. Nucl. Med. 44, 291–315 (2003).
3. K. Wienhard et al., "The ECAT HRRT: performance and first clinical application of the
new high resolution research tomograph," IEEE Trans. Nucl. Sci. 49, 104–110 (2002).
4. C. C. Meltzer et al., "Does cerebral blood flow decline in healthy aging? A PET study
with partial-volume correction," J. Nucl. Med. 41, 1842–1848 (2000).
5. F. J. Beekman, H. W. A. M. de Jong, and S. van Geloven, "Efficient fully 3-D iterative
SPECT reconstruction with Monte Carlo-based scatter compensation," IEEE Trans. Med.
Imaging 21, 867–877 (2002).
6. H. Matsuda et al., "Brain SPET abnormalities in Alzheimer's disease before and after
atrophy correction," Eur. J. Nucl. Med. Mol. Imaging 29, 1502–1505 (2002).
7. G. W. Goerres et al., "Head and neck imaging with PET and PET/CT: artifacts from
dental metallic implants," Eur. J. Nucl. Med. Mol. Imaging 29, 367–370 (2002).
8. F. M. Bengel et al., "Whole-body positron emission tomography in clinical oncology:
Comparison between attenuation-corrected and uncorrected images," Eur. J. Nucl. Med. 24,
1091–1098 (1997).
9. R. de la Fuente-Fernandez et al., "Expectation and dopamine release: Mechanism of the
placebo effect in Parkinson's disease," Science 293, 1164–1166 (2001).
10. O. G. Rousset et al., "Effect of partial volume correction on estimates of the influx and
cerebral metabolism of 6-[(18)F]fluoro-L-dopa studied with PET in normal control and
Parkinson's disease subjects," Synapse 37, 81–89 (2000).
279
CHAPTER 8
Radiation Protection, Standards and

Regulations
8.1. The LNT model is appropriate for the estimation of risk from
low-level (less than 100 mSv/year) radiation, and low levels of radon
in homes should be considered harmful to health
Daniel J. Strom, John R. Cameron and Bernard L. Cohen

Reproduced from Medical Physics, Vol. 25, No. 3, pp. 273-278, March, 1998
OVERVIEW
The linear no-threshold hypothesis is at the heart of radiation risk calculations, standards setting,
and regulatory philosophy. If the LNT ‗‗theory‘‘ is correct, then any small amount of radiation
constitutes a risk to those exposed. On the other hand, if the ‗‗theory‘‘ is wrong, and risks are
much lower than our present regulations are designed to protect against, then we could save
considerable time, effort and expense trying to comply with overly restrictive exposure limits.
Application of the LNT hypothesis has literally devastated at least one industry in the United
States (the nuclear industry) while, at the same time, it has been responsible for spawning others,
such as the home radon-proofing industry. It is also responsible for the employment of large
numbers of regulators, inspectors and, yes, medical and health physicists. This is clearly an
important issue for medical physicists and we are fortunate to have three of the world‘s foremost
experts to debate it in our Point/Counterpoint series.
Arguing FOR both the Motions is Dr. Daniel Strom, Staff Physicist in the Risk Analysis and
Health Protection Group at the Pacific Northwest National Laboratory, Richland, Washington.
Dr. Strom earned his Ph.D. in Environmental Sciences and Engineering at the University of
North Carolina, Chapel Hill in 1983. Among his current research interests are risk analysis, and
protection against radon and its progeny. He claims to be in the middle of the spectrum of views
on dose-response models.
Arguing AGAINST the motion that the LNT model is appropriate for the estimation of risk from
low-level (less than 100 mSv/year) radiation is Professor John Cameron. Dr. Cameron earned his
Ph.D. in Nuclear Physics in 1952 at the University of Wisconsin, Madison, where he has spent
280
almost his entire working life and is now Professor Emeritus in the Department of Medical
Physics. Also throughout his career, Prof. Cameron has been concerned with protecting people
from unnecessary radiation exposure—he ‗‗invented‘‘ the roentgen-area-product concept in
order to protect patients, for example. At the same time, however, he has continually expressed
concern about over stressing the risks of radiation, which he not only considers stifling to
progress (and expensive), but also frightens the general public. He is especially interested in
allaying the fears of the public by educating them about radiation and its effects.
Finally, arguing AGAINST the Motion that low levels of radon in homes should be considered
harmful to health is Professor Bernard Cohen. Dr. Cohen earned his D.Sc. in Physics in 1950 at
the Carnegie Institute of Technology. Since 1958 he has served on the faculty of the University
of Pittsburg, where he is Professor of Physics. He is the author of several books, including A
Homeowners Guide to Radon, as well as numerous publications about radon and radiation.
Professor Cohen is considered one of the world‘s leading experts on the risks associated with
radon in homes.
Argument for both motions: Daniel J. Strom
Opening Statement
The linear, non-threshold (LNT) dose-response model forms the basis for all USA and
international recommendations and regulations for protection of workers and the public from
harmful effects of radiation at low doses. It is not used for high-dose (‗‗deterministic‘‘) effects,
for which nonlinear, threshold models are well established.
The LNT model states that radiation detriment increases as a linear function of dose, without
threshold, when averaged over all ages and both sexes. Detriment is the expectation of harm,
which includes loss of life expectancy or quality of life due to fatal and nonfatal cancers and
heritable ill-health. These are stochastic effects, that is, their frequency in a population, rather
than their severity, is a function of dose.
Radiation protection is a risk management activity. Science is one of many inputs to risk
management. There is no practical way to incorporate everything we know as scientists about
radiation-induced cancer into risk management. We know that radiation biology is at least a 16-
dimensional problem that includes health endpoint, response and projection model, amount of
life lost, portion of organism irradiated, background incidence, who‘s exposed and who‘s
affected, dose, dose rate, dose fractionation, LET (microdosimetry), sex, age at exposure, age at
diagnosis, species, subspecies or genetic predisposition, and other effect modifiers (smoking,
oxygen, diet, etc.), so using only two of these (dose and response) cannot possibly be correct.
While there are clearly human data that show a response threshold for some cancers (bone cancer
from ingested radium, liver cancer from injected thorium, and perhaps lung cancer in
nonsmoking miners exposed to radon progeny), there are many others that show no threshold at
doses of concern in radiation protection (solid tumors in the Japanese bomb survivors, lung
cancer in smoking miners), and one neoplasm, leukemia, for which the dose-response
relationship is significantly nonlinear in the Japanese bomb survivors. There is significant reason
to believe that the mechanisms of carcinogenesis differ for these diseases.
Valid scientific arguments supporting the LNT model include the following: Tumors are of
monoclonal origin; low-dose radiation is a small perturbation in the effect of other carcinogens
281
that have already exceeded most thresholds; miner, bomb survivor, and other human studies for
most cancer endpoints are consistent with LNT; heritable ill-health probably follows LNT, bomb
survivor data are compatible with LNT projections of heritable ill-health from animal studies.
Valid scientific arguments against LNT include the following: some cogent radiation data do not
show LNT behavior for some cancer endpoints; no statistically significant heritable ill-health is
seen in bomb survivors (although this is consistent with the 2 Sv doubling dose from animal
studies). Specious scientific arguments against LNT include the following: ‗‗if you cannot detect
a health effect, it does not exist‘‘; ‗‗if you cannot detect a health effect, it is of no concern‘‘;
bomb survivor and miner studies are ‗‗high dose‘‘ studies that are inappropriately extrapolated to
low doses; oxidative damage is the same for radiation and chemicals; adaptive response occurs;
threshold analogies make sense (e.g., {high, medium, low} applied to {fall, wind, impact});
hormesis is important; some chemical carcinogens have thresholds; energy imparted, not dose, is
the independent variable.
Valid policy arguments for LNT include the following: it errs on the side of safety (it is
‗‗conservative‘‘); it is a politically acceptable status quo; at present, there is no prospect of direct
measurements of effects at doses of interest; a practical system based on LNT has protected
workers. Valid policy arguments against LNT include the following: it has led to expensive risk-
management decisions; optimization has not worked (the ‗‗R‘‘ in ALARA has been ignored);
small lifetime fatal cancer risks may result in insignificant life-shortening. A specious policy
argument for LNT is that a threshold system is impractical.
‗‗All models are wrong, and some are useful‘‘ (Box, 1979). Use of the LNT model as a basis for
setting standards for radiation protection against stochastic effects at low doses still makes good
policy sense. The LNT model should not be used for individual risk predictions (either
prospectively or retrospectively) or for priority-setting; for these applications, the detailed,
unbiased risk assessments that account for all known variables should be used.
Rebuttal to Professor Cameron
Cameron invokes the dose-rate dependency of health effects of radiation, a well-established

phenomenon. It has long been known that if pairs of microscopic DNA lesions are sufficiently
separated either in time or in space, they do not interact; if they occur close enough in time and
space, they may interact.1 This is accounted for in current versions of the LNT model by using a
dose and dose rate effectiveness factor (DDREF).
Three recent reviews2–4 have confirmed that there are thresholds for some kinds of radiation-
induced cancer, and one kind of leukemia has never been seen in excess in irradiated
populations. This in no way implies that there are thresholds for all kinds of cancer, especially
with evidence to the contrary.
Rebuttal to Professor Cohen
Contrary to Cohen‘s claim, there are good data supporting LNT in the dose regions low enough
to be directly applicable to many important radiation protection problems, including indoor
radon. These data are from underground miners, 5,6 indoor case-control studies,7 and Japanese
bomb survivors.8
Cohen‘s free radical argument is irrelevant because oxidative damage by free radicals at single
sites is almost completely repaired, whether the free radicals are caused by chemicals or
282
radiation. The damage of concern from radiation is caused by moderate to large clusters of
ionization9 formed at the end of charged particle tracks (the Bragg peak), for which there is no
chemical analog. Understanding such damage does not require postulating an impairment of
BDM. One-time inductions of adaptive response (‗‗enhancement of BDM‘‘) take significant
doses (e.g., 150 mGy), and like a suntan (also an adaptive response) it fades with a half time of
days to weeks. I know of no evidence that adaptive response can be maintained indefinitely, or
induced by dose rates on the order of 1 mSv per year. It requires no extrapolation from human
data (early radiologists) to conclude that repeated doses of 150 mGy to maintain adaptive
response would cause deterministic effects and excess cancer.
Finally, the county-radon-lung cancer ecologic study is not a logically compelling design.
Conclusions of an ecologic study are good for hypothesis generation, not hypothesis testing. For
‗‗Principles for Evaluating Epidemiologic Data in Regulatory Risk Assessment,‘‘10 see
www.sph.umich.edu/group/eih/UMSCHPS/epidprin.htm.
Concluding remarks
In the face of conflicting science, the LNT model continues to be a useful basis for radiation
protection. It should not be used for individual risk estimates, but it is useful for setting
standards.
REFERENCES
1. A. M. Kellerer and H. H. Rossi, ‗‗The Theory of Dual Radiation Action,‘‘ Current Topics
in Radiat. Res. Quart. 8, 85–158 (1972).
2. R. Cox, C. R. Muirhead, J. W. Stather, A. A. Edwards, and M. P. Little, ‗‗Risk of
Radiation-Induced Cancer at Low Doses and Low Dose Rates for Radiation Protection
Purposes,‘‘ Documents of the NRPB 6, 1–77 (1995).
3. National Council on Radiation Protection and Measurements (NCRP), ‗‗Principles and
Application of Collective Dose in Radiation Protection,‘‘ Report No. 121 (NCRP Publications,
Bethesda, MD, 1995).
4. Advisory Committee on Radiological Protection, ‗‗Biological Effects of Low Doses of
Radiation at Low Dose Rate,‘‘ ACRP-18 (Atomic Energy Control Board of Canada, Ottawa,
Canada, 1996).
5. J. H. Lubin and J. D. Boice, Jr., ‗‗Lung cancer risk from residential radon: meta-analysis
of eight epidemiologic studies,‘‘ J. National Cancer Institute 89, 49–57 (1997).
6. J. H. Lubin, L. Tomásek, C. Edling, R. W. Hornung, G. Howe, E. Kunz, R. A. Kusiak, H.
I. Morrison, E. P. Radford, J. M. Samet, M. Tirmarche, A. Woodward et al., ‗‗Estimating Lung
Cancer Mortality from Residential Radon Using Data for Low Exposures of Miners,‘‘ Radiat.
Res. 147,
126–134 (1997).
7. J. M. Samet, ‗‗Indoor Radon Exposure and Lung Cancer: Risky or Not?—All Over
Again,‘‘ J. National Cancer Institute 89, 4–6 (1997).
8. D. A. Pierce, Y. Shimazu, D. L. Preston, M. Vaeth, and K. Mabuchi, ‗‗Studies of the
Mortality of Atomic Bomb Survivors. Report 12, Part I. Cancer: 1950–1990,‘‘ Radiat. Res. 146,
1–27 (1996).
9. D. T. Goodhead, ‗‗Radiation Tracks in Biological Materials: Initial Damage in Cells,
DNA and Associated Structures,‘‘ in Proceedings of the NCRP No. 13, edited by M. L.
Mendelsohn (NCRP Publications, Bethesda, MD, 1992), pp. 25–37.
283
10. Federal Focus, Inc., Principles for Evaluating Epidemiologic Data in Regulatory Risk
Assessment. Developed by an Expert Panel at a Conference in London, England, October 1995
(Federal Focus, Inc., Washington, DC, 1996).
Argument against the motion that the LNT model is appropriate for the
estimation of risk from low-level (less than 100 mSv/year) radiation: John R.
Cameron
Opening Statement
No! LNT is not appropriate to estimate risks to any population at any dose or dose rate. LNT is
an unrealistic theoretical model contradicted by much human and animal data. Unrealistic
because it is unusual for a biological response to be linear over even one decade. It would be
even more unlikely that it would be linear at or near zero when the body must have natural
defenses to survive the 40 million radioactive disintegrations per hour inside the average adult.
The LNT assumption is alledgedly based on radiation-induced cancer among the A-bomb
survivors which showed an apparent threshold of about 30 cGy, even for leukemia. The dose rate
to the A-bomb victims was about a million times greater than that encountered by radiation
workers. If the dose had been protracted over a few months or years, the apparent threshold
would have been much higher. For example, while A-bomb victims had a high leukemia
incidence eight years post exposure, Chernobyl victims with comparable doses spread over
weeks or months had no significant increase in leukemia. There is good evidence that the body
has a protective mechanism referred to as apoptosis. That is, injured cells are programmed to
‗‗commit suicide‘‘ to protect the organism. Prof. Sohei Kondo 1 calls the low dose rate where all
damaged cells are eliminated the apoptosis dose rate. At the higher necrotic dose rate, apoptosis
cannot keep up. Tissue repair errors lead to cancer induction. Kondo cites two examples to
support his model. Rats exposed to a total of 25 working level (WL) months at rates of 2 WL and
100 WL had markedly different lung cancer incidence. 2 At a rate of 2 WL lung cancer was at the
background rate of about 0.5%. At a rate of 25 WL, a necrotic dose rate, lung cancers were about
three times the background rate. The skin of mice were irradiated with beta rays to a limited area
three times a week for life or until the appearance of skin cancer. 3 At a dose rate of 1.5 Gy/week
there was no skin cancer. At a necrotic dose rate of 3 or more Gy/week there was 100%
incidence of skin cancer. In humans a similar dramatic effect was seen in radium induced
osteogenic sarcomas among the radium dial painters. There was no radium induced bone cancer
until the skeletal dose exceeded 10 Gy (200 Sv). From 20 Gy to 500 Gy the incidence of
osteogenic sarcomas was essentially constant at 2866%. Bond 4 points out that it is inappropriate
to predict individual risks from epidemiological data. He feels radiation is a public health
problem.
Rebuttal to Dr. Strom
When scientists argue it indicates a lack of definitive data. I think all of us agree there are no
definitive data to show radiation risk at the levels now set for radiation workers and the public.
My greatest concern is the use of the LNT model by the news media and others to produce fear.
Many science teachers are often unaware of the relatively large amount of radioactivity in their
own body—almost 10,000 Bq. Our greatest need is to educate the public about radiation. I
suggest that every TV weather map show real time radiation levels in nGy/h for radiation
monitoring stations in their area and around the country. The public would see the actual
radiation levels near nuclear power plants and far from nuclear power plants in the mountains.
284
In addition every commercial passenger airplane should have a clearly visible radiation monitor
that shows the radiation level continously during flight including its value as the plane flies at
high altitude. By this means the public would become familiar with radiation levels and their
variation.
The dose from every diagnostic radiology exam should be explained by the RT in terms of the
time to get approximately the same dose from background radiation. This can be done by means
of a small brochure that gives typical values for common x-ray exams. Medical fluoroscopes
should be required to have a dose-area product meter so that these larger doses can also be
explained in terms of background radiation.
REFERENCES
1. S. Kondo, ‗‗Tissue-repair error model for radiation carcinogenesis,‘‘ Proc. 12th Int.
Congr. Photobiol. (1996).
2. J. P. Morlier et al., ‗‗Lung cancer incidence after exposure of rats to low doses of radon:
Influence of dose rate,‘‘ Radiat. Prot. Dosim. 56, 93–97 (1994).
3. A. Ootsuyama and H. Tanooka, ‗‗One hundred percent tumor induction in mouse skin
after repeated beta induction in a limited dose range,‘‘ Radiat. Res. 115, 486–494 (1988).
4. V. Bond, ‗‗When is a Dose Not a Dose?‘‘ NCRP Lecture 15 (1992).
Argument against the motion that low levels of radon in homes should be
considered harmful: Bernard L. Cohen
Opening Statement
The answer to this question is largely dependent on the validity of the linear no-threshold theory
(LNT) of radiation carcinogenesis. There are no data supporting LNT in the low dose region; it
is based on the following reasoning: Since we believe that even a single particle of radiation
hitting a single DNA molecule in a single cell nucleus can initiate a cancer, the number of
cancers initiated is proportional to the number of such hits, which is proportional to the dose. It
has long been known that there are biological defense mechanisms (BDM) which prevent all but
a very tiny fraction of initiating events from developing into a clinical cancer, but it has been
tacitly assumed that these BDM are not affected by radiation.
It is now recognized1 that cancer initiating hits on DNA molecules, indistinguishable from those
caused by radiation, occur at a very high rate due to random thermal agitation and chemical
attack by free radicals—about 6000 hits per cell each hour, or 50,000,000 per year. Since 1 cGy
(1 rad) of radiation causes only about 20 such hits per cell, it is obvious that the latter are
inconsequential. How, then, can radiation cause cancer? The only possible answer is that
radiation can degrade our BDM. Several biological mechanisms have been proposed to explain
this, but none of them give any reason to believe that this degradation is linearly proportional to
dose, as required to justify LNT.
On the contrary, there is abundant indisputable evidence that low doses of radiation enhance
BDM.2 It has been shown in numerous independent experiments, both in vitro and in vivo, that
low dose pre-exposure substantially reduces the number of chromosome breaks and the number
of gene mutations produced by later high dose radiation exposures. It has also been shown that
285
low dose radiation stimulates the activity of the immune system as measured by various
indicators. Thus, the theoretical basis for LNT is completely negated, and there is a clear
suggestion that low level radiation may actually be protective against cancer.
Experimental data predominantly support the latter viewpoint, or at least the existence of a
threshold below which radiation is essentially harmless. Data on luminous watch dial painters 3
who got radium into their bodies by tipping their brushes with their tongues, shows a clear
statistically significant threshold behavior. Leukemia among Japanese A-bomb survivors,4 and
breast cancer among Canadian tuberculosis patients5 exposed by frequent fluoroscopy, both
show statistically significant decreases with increasing dose in the low dose region.
But the data most directly relevant to our question are from a compilation of average radon levels
in homes for 1729 U.S. counties, well over half of all U.S. counties and comprising about 90% of
the total U.S. population. The results6 show a statistically indisputable tendency for lung cancer
rates, with or without correction for smoking prevalence, to decrease with increasing radon
level; the slope is discrepant with the prediction of LNT by 20 standard deviations! It was shown
that ‗‗the ecological fallacy‘‘ and other weaknesses of ecological studies do not apply to this
work. Effects of over 60 potential confounding factors were studied, several other tests were
applied, and the data and results have been available for two years, but there has been no
explanation for this discrepancy other than that LNT fails in the low dose region, grossly over-
estimating the cancer risk from low level radiation.
Rebuttal to Dr. Strom
Strom states that solid tumors in Japanese A-bomb survivors and lung cancer in miners show no
threshold at doses of concern in radiation protection, and later states that it is ‗‗specious‘‘ to
argue that these ‗‗are high dose studies that are inappropriately extended to low dose.‘‘
The A-bomb survivor data4 show no statistically significant evidence that there is not a threshold
below 25 cSv (25 rem). In fact, using those data, it is easy to show that there is a 30% probability
that the risk decreases with increasing dose up to 20 cSv. By contrast, EPA and NRC are now
squabbling over 0.015 vs 0.025 cSv as a regulatory limit for radiation protection.
If our data showing a strong decrease in lung cancer rates for increasing radon exposures in U.S.
Counties is interpreted directly as risk vs exposure to individuals, there is no statistically
significant discrepancy between it and the miner data. This was shown by Ken Kase at the 1997
Milwaukee meeting of AAPM. These data cover a range of radon levels that EPA estimates to be
causing over 10,000 deaths per year in the U.S.
I do not understand Strom‘s statements that it is specious to argue that ‗‗adaptive response
occurs‘‘ and that ‗‗hormesis is important.‘‘ There is an indisputable body of evidence, accepted
by ICRP and UNSCEAR, supporting adaptive response. There is certainly a great deal of
evidence, albeit not conclusive, supporting hormesis; this has been the topic of several large
international conferences, at least two books, etc. I see nothing specious about pointing out that
adaptive response or hormesis can explain why linear no-threshold theory fails.
I take this opportunity to apologize for my misinterpretation of the Billen paper. 1 The ‗‗6000 hits
per hour‘‘ are not necessarily cancer initiating events, and there are certainly differences between
the damage done by radiation and by chemical attack. The situation is quite complicated, but
Billen‘s paper concludes that ‗‗spontaneous DNA damage may be many orders of magnitude
286
greater than that caused by low radiation doses.‘‘ Billen now says that the ‗‗may be‘‘ was over-
conservative and can be modified to ‗‗according to available evidence, should be.‘‘
REFERENCES
1. D. Billen, Radiat. Res. 124, 242–245 (1990).

2. UNSCEAR 1994 Report, Appendix B.
3. R. E. Rowland, A. F. Stehney, and H. F. Lucas, Radiat. Res. 76, 368–383 (1978).
4. D. A. Pierce et al., Radiat. Res. 146, 1–27 (1996).
5. A. B. Miller et al., N. Engl. J. Med. 321, 1285–1289 (1989).
6. B. L. Cohen, Health Phys. 68, 157–174 (1995).
287
8.2. Exposure to residential radon causes lung cancer
R. William Field and Philippe J. Duport

OVERVIEW
In many geographic regions, exposure to radon contributes significantly to background radiation

levels. According to the linear, no-threshold model, this contribution raises the cancer risk of
exposed individuals. Radon levels in homes can be reduced, but often the costs are substantial.
Some experts feel that the costs are not justifiable, because epidemiological studies have failed to
show a correlation between radon levels and cancer incidence. This debate between experts
creates public uncertainty and frustration. It is this debate that is the subject of this month's
point/counterpoint.
Arguing for the Proposition is R. William Field, Ph.D. Dr. Field is a Cancer Epidemiologist in
the Department of Epidemiology, an adjunct professor in the Department of Occupational and
Environmental Health, and a member of the graduate faculty in the College of Public Health at
the University of Iowa. He received his Doctorate Degree in Preventive Medicine from the
University of Iowa in 1994 and since that time has published over 40 articles, including book
chapters, related specifically to radon and environmental radiation. Dr. Field has expertise and
extensive work experience in Health Physics, Environmental Health, and Epidemiology. He has
served as associate editor for the following journals, Reviews on Environmental Health, Health
Physics, and Journal of Toxicology and Environmental Health.
Arguing against the Proposition is Philippe J. Duport, Ph.D. Dr. Duport is Founder and Director
of the International Centre for Low Dose Radiation Research at the University of Ottawa. Dr.
Duport has thirty years of field experience in aerosol physics as well as in environmental and
radiation protection in the uranium industry in Canada and France. He became Head of Health
and Environmental Effects Research Section of the Atomic Energy Control Board of Canada. He
was also a member of the AECB delegation to the United Nations Scientific Committee on the
Effects of Atomic Radiation (UNSCEAR) and to OECD-NEA specialized working groups on
radon and thoron dosimetry.
FOR THE PROPOSITION: R. William Field, Ph.D.
Opening Statement
Radon decay products (radon) are well established lung carcinogens.1,2 Over 20 occupational
epidemiologic studies of radon-exposed underground miners have unequivocally demonstrated
that prolonged exposure to radon increases the risk of lung cancer. 3 Pooled studies of 65 000
miners found a linear relationship between radon exposure and lung cancer deaths.3,4 This
relationship was maintained, even among a subgroup of miners that had lower exposures
extending into the range for some homeowners.3,5 These findings suggest that residential radon
also carries a risk of cancer.3,4,5 The National Academy of Sciences (NAS) has developed radon
288
risk models based on miner studies that project 18 600 lung cancer deaths each year in the United
States caused by residential radon exposure. 3
There is mounting scientific evidence to support the argument that residential radon exposure
causes lung cancer. A 1997 National Cancer Institute (NCI) meta-analysis6 examined the effects
of residential radon exposure on lung cancer risks, using data from eight of the previously-
published large scale residential case-control studies performed independently in Canada, China,
Finland, Sweden, and the United States. The results reveal an excess risk of cancer of 14% at 148
Bq.m–3. Subsequent major residential case-control studies in Europe7,8 and China,9 as well as
preliminary findings from the pooling10 of North American residential radon studies, which
includes 4 081 cases and 5 281 controls, closely agree with the NAS projected risk estimates.
Moreover, two recent residential case-control studies in Missouri11 and Iowa,12 that used
enhanced dosimetry methods, found even higher risk estimates. These findings strongly suggest
that previous radon studies may have actually underestimated the risk posed by residential radon,
because exposure misclassification was found to bias the studies toward finding no association. 13
In summary, risk estimates from rigorously designed analytic epidemiologic studies provide
compelling evidence that prolonged residential radon exposure increases the risk of lung cancer.
Rebuttal
In theory, a single alpha particle can trigger double strand DNA breaks, leading to cancerous
transformation of a single cell.3 In vitro studies show that cells exposed to high-LET radiation
send out "signals," which cause chromosomal damage and transformation to nearby unirradiated
"bystander" cells as well.14,15 Thus, based on the radiobiological effects of high-LET radiation,3,16
low-dose radon exposure can induce cancer.
There are several reasons why one should not use animal studies to address the risk of low-dose
radon exposure on the development of lung cancer in humans. First, rats do not develop small
cell carcinoma, a common type of lung cancer in people. Second, low dose rate studies in animals
may underestimate risk, because the time needed for the development of cancer may exceed the
lifetime of the animal. Finally, one cannot extrapolate results species-to-species, because there
are differences in dose-response relationships among species. Therefore, the best studies to assess
the risk of radon-induced cancer in humans are the uranium miner studies and the rigorously
designed residential radon case-control epidemiologic studies.
Duport offers two reasons why indoor radon risk cannot be extrapolated from uranium miner
studies. His first reason, that the uranium miners would have been exposed to other sources of
radiation, is moot, because lung cancer risk from radon exposure is similar between uranium and
nonuranium miners. His second reason, that nonsmoking miners exposed to radon doses less than
400 WLM did not develop lung cancer, deserves clarification. In fact, there were cases of lung
cancer in the exposure categories of nonsmoking miners below 400 WLM. The estimated excess
relative risk per WLM was higher for never-smoking miners than for smoking miners.3 This
agrees with data from residential radon studies, which indicate that residential radon exposure
increases lung cancer risk even in nonsmokers.10,17
Duport dismisses the epidemiologic studies of residential radon exposure as confusing. The
residential radon epidemiologic studies have generated conflicting findings. Ecologic radon
studies are the principle source of confusion in radon epidemiology. 18,19,20 Epidemiologists
maintain that the ecologic study design should be reserved for generating hypotheses rather than
289
estimating risk. Our research group13 and others21 have argued that the lack of significant
findings of some of the earlier residential case-control studies is attributable to random
misclassification of risk factors, primarily from poor assessment of radon exposure, which
reduces a study's power to detect an association.
Duport suggests that epidemiologic studies would be more credible if confidence intervals
included ALL dosimetric errors and uncertainties. Changing the way confidence intervals are
calculated in epidemiology will not change the central estimate of risk. Nonetheless, Duport's
statement indirectly addresses the importance of accurate dosimetry in environmental
epidemiology, a topic of great priority to our research group.12,13,22 In fact, substituting less
accurate dosimetry data in case-control studies reduces the ability to detect an
association.10,11,12,13,21,22 Furthermore, studies that incorporate enhanced dosimetry methods find
higher risk estimates.10,11,12,13
In summary, an abundance of scientific evidence has clearly demonstrated that prolonged

residential radon exposure increases the risk of lung cancer.
AGAINST THE PROPOSITION: Philippe J. Duport, Ph.D.
Opening Statement
The absence of an effect can never be proven with absolute certainty. Nevertheless, there are
experimental, epidemiological and dosimetric reasons to doubt that indoor radon decay products
(RDP) cause lung cancer, at least at concentrations below several hundreds of Bq/m3.
Experiment
It has been shown that rats exposed to a low dose [25 working-level months (WLM)] of RDP at
high dose rates of 150 and 100 WL 6×105 and 4×105 Bq/m3 show a significant excess of lung
cancers. Paradoxically, the same dose given at 2 WL ( 8 000 Bq/m3) has a nonsignificant
23
protective effect. This is consistent with other unambiguous thresholds, at doses up to several
grays of radiation, in animals24,25 and in humans.26,27 Only an radiation weighting factor
wR( ) 0 at low dose rates explains the absence of risk.
Epidemiology
1. Wrong assumptions in uranium miner studies.
The risk of cancer in an organ is, theoretically, proportional to the total organ dose. In uranium
mines, doses other than from RDP are not negligible. In addition, with wR( ) 1 the relevant
lung dose for risk estimation is the total absorbed dose. In uranium miner cohorts for which
individual doses from each radiation source are available and reliable, the RDP risk is
overestimated by a factor of 2 to 4, or more if doses received in "neglected" mines and dose
misclassification are also considered. 28 The corrected excess relative risk per working level
month (ERR/WLM) is close to that of Chinese tin miners (0.001/WLM), 29 which increases the
likelihood of no effect at the lowest exposures.
290
2. Lung cancer in nonsmoking uranium miners.
The lowest exposure in non-smoking uranium miners with lung cancer is about 450 WLM. 30
3. Confusing indoor studies.
Upon visual inspection of the data points evenly and widely distributed about the no-effect line,
what confidence can be granted to a positive trend in a meta-analysis of indoor RDP,6 when well-
designed studies support either the LNT, 31 a U-shaped response32 or no effect whatsoever?33
Dosimetry
Risk is proportional to dose. RDP dosimetry is very uncertain;28,34 uncertainties in RDP lung dose
are very large and impossible to quantify. RDP epidemiologic studies would be more credible
(but would they still appear meaningful?) if confidence intervals included ALL dosimetric errors
and uncertainties, in addition to statistical mortality uncertainty. In other sciences, peer reviewers
would challenge papers in which error bars take only a fraction of all possible errors into
account.
Conclusions
(1) Animal and human studies show, convincingly, that low doses and dose rates of alpha
radiation have no health effects on the lung or other organs. Biological arguments should be
offered to explain why such effects should exist for indoor RDP alpha radiation.
(2) Indoor radon risk cannot be extrapolated from biased miner studies.
(3) Currently published epidemiologic radon studies give a false sense of accuracy because their
confidence intervals, as large as they may be, are artificially narrow: They take into account only
the quantifiable (and arguably the smallest) part of all errors and uncertainties.
Rebuttal
Dr. Field's arguments are based solely on an epidemiologic construct that relates risk to radon
concentration. However, it is the dose (or exposure), not the concentration, that determines the
risk. The dose and the concentration are not rigorously related for indoor radon. Radon risks at
low doses and dose rates, as extrapolated from uranium miner studies, are overestimated. This is
because significant lung doses from gamma radiation, inhaled ore dust and doses received in
"neglected" mines have been systematically ignored. In individual and pooled miner studies,
confidence intervals at low exposures accommodate any dose-response shape, including the
presence of thresholds. These confidence intervals are also underestimated, and would be much
larger if all dosimetry and mortality errors were taken into account.
Dr. Field cites European, Chinese and US indoor case-control studies to support his argument.
However, other case-control studies32,33 tend to support no-effect or threshold arguments. Why
does my colleague ignore these studies? With all errors taken into account, the number of
projected radon-induced lung cancers in the US would be between zero and a high upper limit.
291
Low doses and dose rates of alpha radiation, including RDP, are probably not carcinogenic in
humans26,27,29 or in animals.23,24,25 Dr. Field does not explain why RDP alpha emitters are
carcinogenic while other alpha emitters are not. He does not address why the observed
ineffectiveness of alpha radiation in humans and animals should be neglected in the interpretation
of indoor radon risk. In brief, Dr. Field's arguments would be stronger if they reconciled indoor
radon epidemiology (all doses and errors taken into account) with dosimetry and indisputable
human and animal data at low dose rates of alpha radiation (including RDP). The
noncarcinogenicity of low doses and dose rates of alpha radiation (including RDP) is established
more rigorously than the carcinogenicity of indoor radon.35
REFERENCES
1. ATSDR, Agency for Toxic Substances and Disease Registry, "Toxicological profile for
radon," Final Report, ATSDR/TP-90/23 (Atlanta, GA, ATSDR). Public Health Service, U.S.
Department of Health and Human Services, NTIS Accession No. PB91-180422, 1990.
2. IARC, International Agency for Research on Cancer, "IARC monographs on the
evaluation of carcinogenic risks to humans," Man-made Mineral Fibres and Radon, Vol. 43
(Lyon, France, 1988).
3. National Research Council, Health Effects of Exposure to Radon, BEIR VI, Committee on
Health Risks of Exposure to Radon (BEIR VI), Board on Radiation Effects Research,
Commission on Life Sciences (National Academy, Washington, DC, 1998).
4. J. H. Lubin et al., "Lung cancer in radon-exposed miners and estimation of risk from
indoor exposure," JNCI, J. Natl. Cancer Inst. 87, 817–827 (1995).
5. J. H. Lubin et al., "Estimating lung cancer mortality from residential radon using data for
low exposures of miners," Radiat. Res. 147, 126–134 (1997).
6. J. Lubin and J. Boice, Jr., "Lung cancer risk from residential radon: Meta-analysis of
eight epidemiologic studies," JNCI, J. Natl. Cancer Inst. 89, 49–57 (1997).
7. L. Kreienbrock, M. Kreuzer, M. Gerken, G. Dingerkus, J. Wellmann, G. Keller, and H. E.
Wichmann, "Case-control study on lung cancer and residential radon in western Germany," Am.
J. Epidemiol. 153, 42–52 (2001).
8. S. Darby, E. Whitley, P. Silcocks, B. Thakrar, M. Green, P. Lomas, J. Miles, G. Reeves,
T. Fearn, and R. Doll, "Risk of lung cancer associated with residential radon exposure in South-
West England: a case-control study," Br. J. Cancer 78, 394–408 (1998).
9. Z. Wang et al., "Residential radon and lung cancer risk in a high-exposure area of Gansu
Province, China," Am. J. Epidemiol. 155, 554–564 (2002).
10. D. Krewski et al., "A combined analysis of North American studies of lung cancer and
residential radon exposures," American Statistical Association Conference on Radiation and
Health, Deerfield Beach, Florida, 2002.
11. M. C. Alavanja, J. H. Lubin, J. A. Mahaffey, and R. C. Brownson, "Residential radon
exposure and risk of lung cancer in Missouri," Am. J. Public Health 89, 1042–1048 (1999).
12. R. W. Field, D. J. Steck, B. J. Smith, C. P. Brus, J. S. Neuberger, E. L. Fisher, C. E. Platz,
R. A. Robinson, R. F. Woolson, and C. F. Lynch, "Residential radon gas exposure and lung
cancer: The Iowa radon lung cancer study," Am. J. Epidemiol. 151, 1091–1102 (2000).
13. R. W. Field, B. J. Smith, D. J. Steck, and C. F. Lynch, "Residential radon exposure and
lung cancer: Variation in risk estimates using alternative exposure scenarios," J. Expo. Anal.
Environ. Epidemiol. 12, 197–203 (2002).
14. S. Sawant, G. Randers-Pehrson, C. R. Geard, D. J. Brenner, and E. J. Hall, "The
bystander effect in radiation oncogenesis, I. Transformation in C3H10T½cells in vitro can be
activated in the unirradiated neighbors of irradiated cells," Radiat. Res. 155, 397–401 (2001).
292
15. D. J. Brenner and R. K. Sachs, "Do low dose-rate bystander effects influence domestic
radon risks?" Int. J. Radiat. Biol. 78, 593–604 (2002).
16. NCRP, Evaluation of the Linear-Nonthreshold Dose-Response Model for Ionizing
Radiation, Recommendations of the National Council on Radiation Protection and
Measurements, NCRP Report No. 136 (National Council on Radiation Protection and
Measurements, Bethesda, MD, 2002).
17. M. Kreuzer, M. Gerken, L. Kreienbrock, J. Wellmann, and H. E. Wichmann, "Lung
cancer in lifetime nonsmoking men—Results of a case-control study in Germany," Br. J. Cancer
84, 134–140 (2001).
18. B. J. Smith, R. W. Field, and C. F. Lynch, "Residential 222Rn exposure and lung cancer:
Testing the linear no-threshold theory with ecologic data," Health Phys. 75, 11–17 (1998).
19. J. H. Lubin, "The potential for bias in Cohen's ecological analysis of lung cancer and
residential radon," J. Soc. Radiol. Prot. 22, 141–148 (2002).
20. J. Puskin, "Smoking as a confounder in ecologic correlations of cancer mortality rates
with average county radon levels," Health Physics 84, 526–532 (2003).
21. M. C. Alavanja, J. H. Lubin, J. A. Mahaffey, and R. C. Brownson, "Re: Residential radon
gas exposure and lung cancer: The Iowa radon lung cancer study (comment)," Am. J. Epidemiol.
152, 895–896 (2000).
22. R. W. Field, D. J. Steck, C. F. Lynch, C. P. Brus, J. S. Neuberger, and B. C. Kross,
"Residential radon-222 exposure and lung cancer: exposure assessment methodology," J. Expo.
Anal. Environ. Epidemiol. 6, 181–195 (1996).
23. Monchaux G. Contribution of animal experimental data for the risk assessment of
exposure to radon decay products, in Proceedings of Natural Radiation Environment VII
Conference (NRE VII), edited by J. P. McLaughlin, F. Steinhausler, and S. E. Simopoulos,
International Congress Series, Rhodes, Greece, 20–24 May 2002 (Elsevier Science B. V.,
Amsterdam, 2002).
24. C. L. Sanders, K. E. McDonald, and J. A. Mahafey, "Lung tumor response to inhaled Pu
and its implications for radiation protection," Health Phys. 55, 455–462 (1988).
25. R. G. White, O. G. Raabe, M. R. Culbertson, N. S. Parks, S. J. Samuels, and L. S.
Rosenblatt, "Bone sarcoma characteristics and distribution in Beagles injected with radium 226,"
Radiat. Res. 137, 361–370 (1994).
26. R. D. Evans, "The effects of skeletally deposited Alpha-ray emitters in man," Br. J.
Radiol. 39, 881–895 (1967).
27. M. Andersson and H. H. Strom, "Cancer incidence among Danish thorotrast patients,"
JNCI, J. Natl. Cancer Inst. 4, 1318–1325 (1992).
28. P. Duport, "Is radon risk overestimated? Neglected doses in the estimation of the risk of
lung cancer in uranium underground mines," Radiat. Prot. Dosim. 98, 329–338 (2002).
29. R. J. Roscoe, K. Steenland, and W. E. Halperin, "Lung cancer mortality among
nonsmoking uranium miners exposed to radon daughters," J. Am. Med. Assoc. 262, 629–633
(1989).
30. J. H. Lubin, J. D. Boice, Jr., C. Edling, R. W. Hornung, G. Howe, E. Kunz, R. A. Kusiak,
H. I. Morrisson, E. P. Radford, J. M. Samet, M. Tirmarche, A. Woodward, Y. S. Xiang, and D.
A. Pierce, Radon and Lung Cancer Risk: A Joint Anaysis of 11 Underground Miners Studies,
U.S. Department of Health and Human Services, Public Health Service, National Institutes of
Health, NIH Publication No. 94-3644 (1994).
31. R. W. Field, D. J. Steck, B. J. Smith, C. P. Brus, J. S. Neuberger, E. F. Fisher, C. E. Platz,
R. A. Robinson, R. F. Woolson, and C. F. Lynch, "Residential radon gas exposure and lung
cancer: The Iowa radon lung cancer study," Am. J. Epidemiol. 151, 1091–1102 (2000).
293
32. J. Conrady, K. Martin, A. Poffijn, M. Tirmarche, J. Lembecke, D. M. Thai, and H.

Martin, High residential radon health effects in Saxony (Schneeberg study), Report Contract No.
FI4P-CT95-0027, European Community, 1999 (unpublished).
33. E. G. Létourneau, D. Krewski, N. W. Choi, M. G. Goddard, R. J. McGregor, J. M.
Zielinski, and J. Du, "Case-control study of residential radon and lung cancer in Winnipeg,
Manitoba, Canada," Am. J. Epidemiol. 140, 310–322 (1994).
34. A. Birchall and A. C. James, "Uncertainty in analysis of the effective dose per unit
exposure from 222Rn D.P. and implications for ICRP risk-weighting factors," Radiat. Prot.
Dosim. 53, 133–140 (1994).
35. G. Taubes, "Epidemiology faces its limits," Science 269, 164–169 (1995).
294
8.3. A prospective study should be performed to test the hypothesis

that an increase in background radiation to residents in the gulf
states will increase their longevity
John R. Cameron and Jeffrey Kahn
OVERVIEW
Radiation hormesis suggests that small exposures to radiation over extended periods improve
longevity by reducing the risk of disease. One can envision various prospective studies to test the
concept of radiation hormesis. The limitation of these studies is that they may be considered
unethical. One such study is proposed in this issue of Point/Counterpoint.
Arguing for the Proposition is John Cameron, Ph.D. John Cameron completed his Ph.D. from
UW—Madison in 1952 in nuclear physics and switched to medical physics in 1958. He was a
founding member of the AAPM and its tenth president. His research interests were in TLD,
accurate bone measurement and instrumentation for QC of x-ray images. He co-authored several
books. In 1980 he was the founding chair of the Department of Medical Physics at UW—
Madison. Since his retirement in 1986 he has sought to educate the public about radiation. He
believes that low dose rate radiation is beneficial and that human research is necessary to
determine the optimum dose rate.
Arguing against the Proposition is Jeffrey Kahn, Ph.D., M.P.H. Dr. Kahn is Director of the
Center for Bioethics, and Professor of Medicine, University of Minnesota. From April 1994 to
October 1995 Dr. Kahn was Associate Director of the White House Advisory Committee on
Human Radiation Experiments. Dr. Kahn works in a variety of areas of bioethics, exploring the
intersection of ethics and public health policy, including research ethics; ethics and genetics; and
ethical issues in public health. He has published over 50 articles in the medical and bioethics
literature, and his book credits include Beyond Consent: Seeking Justice in Research (Oxford
University Press, 1998). He also writes the bi-weekly bioethics column "Ethics Matters" for
CNN.com.
FOR THE PROPOSITION: John Cameron, Ph.D.
Opening Statement
Arguments in science indicate a lack of convincing data. Probably no other aspect of radiation
protection has been as contentious as the present assumption that even the smallest amount of
radiation may cause an increase in cancer. The cost to society of this assumption is staggering. In
addition, it has contributed significantly to worldwide radiation phobia. Despite the lack of human
data to support the assumption at typical background dose rates, the present policy has been
recently iterated once again.1 It seems likely that the only way to resolve the controversy is a
human radiation study, such as that stated in the proposition. 2 Further, I believe that such a study
would be entirely ethical.
295
A study by Jagger compared cancer mortality and background radiation rates in three mountain
states with those in three gulf states. 3 The mountain states had annual background rates about three
times greater than the gulf states. However, the cancer mortality in the gulf states was about 25%
greater than in the mountain states. This suggests that people in the gulf states are suffering from
higher cancer rates due to a radiation deficiency. If the proposed study is unethical then it is
equally unethical not to warn people that they should not live in a mountain state because of the
increased radiation. Nobody warns the public that flying in jet planes increases the radiation dose.
More convincing data on health benefits of increased radiation comes from several epidemiological
studies. The best radiation worker study is the nuclear shipyard workers study (NSWS) where
cohort and controls were identical except for the radiation exposure to the cohort.4 In this study the
health of 28 000 nuclear workers with the highest cumulative doses (>5 mSv) was compared with
the health of 32 500 age-matched and job-matched unexposed shipyard workers. The cancer
mortality of the nuclear workers was four standard deviations lower than that of the unexposed
workers, and the death from all causes of the nuclear workers was 16 standard deviations (P<10–16)
lower. The annual dose to the nuclear workers was comparable to that received by residents in the
mountain states.
A study of British radiologists from 1897 to 1997 shows a similar health improvement. 5 The high
occupational exposures of early radiologists (1897–1920), with annual doses estimated at 1 Gy,
were associated with a cancer death rate 75% greater than that of all male physicians in England
and Wales. However, their non-cancer death rate was 14% lower than the control group. The death
rate from all causes was slightly less than the control group. That is, despite their high doses,
radiologists had no loss of longevity. British radiologists who first joined a radiological society in
1955–1979 had a non-cancer death rate 36% lower than the control group, and their death rate from
all causes was 32% lower (P<0.001). Their annual dose was estimated to be about 5 mSv.
The evidence suggests that there is no increased risk of cancer at low doses of radiation. That is
why it is ethical to conduct the experiment suggested in the proposition. The results might settle the
low-dose radiation risk issue once and for all.
Rebuttal
I agree with much of Dr. Kahn's statement but I strongly disagree with his assumption that there is
an "unacceptable level of risk" at the dose rate found in the mountain states where millions of
people live longer lives with less cancer than people in the gulf states. The proposed increase in
radiation is about 3% of the dose rate to some people who live in Ramsar, Iran with no obvious
increase in cancer mortality. I would be happy to participate in such a study although I would
prefer a dose rate of 0.1 Gy/y. It seems odd for Dr. Kahn to accept the LNT assumption as a
scientific fact while ignoring the better health of the British radiologists and the nuclear shipyard
workers.
AGAINST THE PROPOSITION: Jeffrey Kahn, Ph.D., M.P.H.
Opening Statement
The process for human subject research in the U.S. is predicated on the principle of protection,
against which any proposed research must be evaluated. This principle evolved out of a history of
exploitation of human subjects, shortcomings in questionable or nonexistent informed consents,
296
and research that posed significant risk without the potential for direct medical benefit to the
subjects. In response to this history, federal rules were created that require prospective review of
any research involving human subjects. This review must assure that proper informed consent
occurs, and take into account whether the level of risk in the research is acceptable, whether the
potential benefit of the research sufficiently offsets the risks entailed by it, and most important,
whether the distribution of the risks and benefits from the research is acceptable. So it is not
enough for risky research to be justified by the potential benefit that it will yield. It is also
necessary that the potential benefit of such research accrues to the subjects accepting the risk.
For example, it would be unethical to test a new form of chemotherapy on otherwise healthy
subjects to better understand its toxicity, even though one could argue that the significant harm
posed to the subjects would be balanced by the benefit of the information gained. The problem is
that the distribution of risks is unfair—all the risks would be born by the subjects, while all the
benefits would accrue to others. Similarly, this Point/Counterpoint raises the question of whether
research on radiation hormesis produces risks without benefits to those exposed to the research, or,
if there are benefits, whether they are sufficient to offset the risks.
The problem with this question is the uncertainty of the risks posed by such research. This
uncertainty affects not only the ability to evaluate the risk-benefit balance of the research, but also
the capacity to assess the issue of risk distribution. What risks would researchers disclose to
potential subjects of such research, and how certain could they be in their disclosure? Without
sufficient information about both the risks and benefits of the research, informed consent is
impossible. Even if we had more complete information, it is questionable that prospective subjects
would receive any disclosure in radiation hormesis research as objective information. Our country's
history of intentional research-related radiation exposures and environmental releases, and the
ongoing health effects claimed by those exposed, have created an atmosphere of suspicion and
distrust with regard to radiation exposures.
Then how might such research proceed? More sophisticated studies of epidemiological data
regarding long-term low-dose exposure—potentially of those exposed occupationally—could
provide much-needed information and guidance for future studies in humans. If human subjects
were intentionally exposed to low-dose radiation, their health would need to be monitored very
carefully and frequently to identify and treat negative health effects quickly.
One measure of the acceptability of any controversial research proposal is whether researchers
would participate themselves or enroll their children. The answer is often instructive if not
definitive. There are likely many research questions that we'd like to be able to answer, but they
entail unacceptable levels or distributions of risk, or carry too much uncertainty. For these research
questions, the price of the answers is simply too high.
Rebuttal
Research that attempts to assess the health effects of environmental exposures generally proceeds
from the assumption that increased rates of exposure bring increased risk of harm. That's why
controlled experiments in which otherwise healthy research subjects are dosed with a presumed
toxin raise red flags—they pose risk to subjects without offsetting direct potential benefit to them.
But the title of Prof. Cameron's statement proposes a study to assess the human health benefits of
low-level radiation exposure, which would be quite a different matter. Offering subjects the
opportunity to participate in research that is intended to benefit their health raises far different
297
ethical concerns—not about protection, but about who gets to participate, and how we can ensure
that they are equitably selected.
The dilemma of this Point/Counterpoint topic is in the suggestion that participating in radiation
exposure research offers health benefits. There is overwhelming evidence that higher doses of
radiation increase cancer risk, so why not start from the presumption that low doses pose risk, too,
rather than the counterintuitive hypothesis that it will be beneficial? Why does this distinction
matter? It is critical both for how we think about the risk–benefit ratio of the proposed research,
and more importantly, for how prospective subjects would perceive the risks and benefits of their
participation.
Even the text that follows the title of Prof. Cameron's statement proceeds from the presumption that
such exposure carries risk. He suggests that the only way to resolve the controversy over the
"assumption that even the smallest amount of radiation may cause an increase in cancer" is to
perform a controlled prospective study of the health effects of intentional exposure to small
amounts of radiation. So the proposal is not to assess the benefit, but the risks of low-level
radiation. Can we conclude from the range of epidemiological studies offered by Prof. Cameron
that low level radiation exposure is actually healthful? Such data cannot show cause and effect—
any claim that it did would be to commit the epidemiological fallacy. Before we can begin to
measure the potential health benefits of radiation exposure in otherwise healthy subjects, we need
to be confident that exposure carries minimal risks of harm. But in an era of increasing protection
of human subjects, it will be difficult to perform research with risk profiles that are unknown at
best, and pose significant risks of harm at worst.
REFERENCES
1. NCRP Report No. 136, Evaluation of the Linear-Nonthreshold Dose-Response Model for
Ionizing Radiation, National Council for Radiation Protection and Measurement, Bethesda, MD,
2001.
2. J. R. Cameron, "Is radiation an essential trace energy?" Physics and Society, October
2001 (also available from http://www.aps.org/units/fps/oct01/a5oct01.html).
3. J. Jagger, "Natural background radiation and cancer death in Rocky Mountain and gulf
coast states," Health Phys. October, 428–434 (1998).
4. G. Matanoski, "Health effects of low-level radiation in shipyard workers final report,"
Baltimore, MD, DOE DE-AC02-79 EV10095 (1991).
5. A. Berrington, S. C. Darby, H. A. Weiss, and R. Doll, "100 years of observation on
British radiologists: mortality from cancer and other causes 1897–1997," Br. J. Radiol. 74, 507–
519 (2001).
298
8.4. Radiation hormesis should be elevated to a position of scientific

respectability
John R. Cameron and John E. Moulder
Reproduced from Medical Physics, Vol. 25, No. 8, pp. 1407-1410, August, 1998
OVERVIEW
Hormesis, the pharmaceutical principal that ‗‗A weak stimulus might stimulate what the same,
but stronger, stimulus inhibits‘‘ is traceable to ancient times. All substances exhibit toxic effects
at the wrong doses. But virtually all substances exhibit either no or beneficial effects at other
doses. Since the discovery of x rays in 1895, many articles have been published demonstrating a
hormetic effect of ionizing radiation at low doses. Yet the no-threshold dose-effect model of
radiation injury ignores these articles and assumes that any exposure to ionizing radiation, no
matter how small, is potentially harmful to human health. This assumption has been challenged
over the years and with increased intensity recently, in part because advocates of radiation
hormesis insist that the effect is scientifically credible. Others insist that the effect is not credible,
in part because no biological model exists to explain radiation hormesis. The issue is
controversial and important. This edition of Point/Counterpoint addresses the controversy.
Arguing for the proposition is John R. Cameron. Professor Cameron spent most of his career
(1958–1986) in Medical Physics at the University of Wisconsin-Madison. In the early 60‘s he
helped develop thermoluminescent dosimetry (TLD) and pioneered the photon absorptiometry
method of bone mineral measurement. In the 70‘s he promoted better quality control of x-ray
imaging and gathered distinguished group of medical physicists at UW. In 1981 he was the
founding Chair of the Department of Medical Physics at UW. Since his retirement in 1986 he has
devoted much time to education about the lack of risk and possible benefit from small doses of
radiation.
Arguing against the proposition is John E. Moulder. Dr. Moulder received his Ph.D. in Biology
from Yale University in 1972. Since 1978, he has served on the faculty of the Medical College of
Wisconsin, where he is Director of the Radiation Biology Program. His primary research interest
is the biological basis for carcinogenesis and cancer therapy. He has published extensively in this
area, and has served on the Experimental Therapeutics and Radiation Review Groups for the
U.S. National Institutes of Health. Dr. Moulder is on the Editorial Board of Radiation Research,
and is an elected member of the Committee on Man and Radiation of the IEEE. He has also
served on the Wisconsin Radiation Protection Council, and on state and local advisory groups
concerned with environmental health, pesticides and nonionizing radiation. Dr. Moulder is
actively involved in educating the public on realistic assessment of cancer risks.
For the proposition: John R Cameron
Opening Statement
Many radiation scientists have always believed that radiation hormesis is scientifically
respectable. The lack of research funding for radiation hormesis is because it is not ‗‗politically
299
respectable‘‘—that is, it contradicts the assumption that radiation risk extends linearly to zero
dose.
Hormesis—health benefits from small quantities of a toxic substance—is well accepted in

medical practice. Obvious examples are medications, hormones, and the 17 essential trace
elements. All are poisonous in large amounts but beneficial in small quantities. The best
collection of references—over 1300—on radiation hormesis appear in two books by T. D.
Luckey: Hormesis with Ionizing Radiation (CRC Press, Boca Raton, 1981) and Radiation
Hormesis (CRC Press, Boca Raton, 1991). An excellent recent review by Luckey appeared in
21st Century, Fall 1996.
Evidence for radiation hormesis appears in NCRP Report No. 104 (1990). On p. 118 it states:
‗‗Maisin et al. (1983) have reported a significant decrease in lung carcinomas after exposure to
0.02 Gy of 23 MeV neutrons.‘‘ On p. 119, Fig. 6.10 shows that gamma radiation reduces the
incidence of lung adenomas from about 30% for the controls to about 20% at a dose of about
0.25 Gy. Later, on the same page: ‗‗In BALB/c mice, a decreased incidence was found for
neutrons in the range between about 0.05 to 0.2 Gy dose levels...‘‘ Data in Miller et al. [New
Eng. J. Med. 321, 1285–1289 (1989)] show that breast cancer mortality decreased to 66% of
controls (p,0.05) for cumulative fluoroscopic exposures of 10–19 cGy. (See also, Pollycove in
Physics & Society News, pp. 6–8, April 1998.)
While radiation hormesis plays a role in reducing cancer its primary benefit to the public is in
improved health, probably through a stimulated immune response. Two examples: (1) Japanese
A-bomb survivors, despite about 400 radiation induced cancer deaths, are living longer on the
average than the unexposed controls. (2) The Nuclear Shipyard Worker Study (Matanoski DOE
report, 1991) shows that the 29 000 nuclear shipyard workers with the highest cumulative doses
were much healthier than the 33 000 age and job matched workers on non-nuclear ships. The
cancer death rate was lowest for the nuclear workers but the really significant result was a 24%
(16 std. dev.) lower death rate from all causes compared to the controls. This important study,
completed in 1988, and reported in UNSCEAR 1994 has yet to be published in a scientific
journal.
Strong scientific support for radiation hormesis comes from cellular studies. Feinengdegen et al.
‗‗Low level radiation may protect against cancer‘‘ (Physics & Society News, pp. 4–6, April
1998) present data from studies of rodents and humans to show that low level radiation (<0.2
Gy) is beneficial to health. (See Fig. 1.) The article documents four beneficial effects: (1)
damage prevention by temporarily stimulated detoxification of molecular radical species; (2)
damage repair from temporary stimulation of repair mechanisms; (3) damage removal by
apoptosis which results in cell death in response mainly to DNA alterations; and (4) damage
removal by stimulating the immune response. While doses in the range of 0.1 to 0.2 Gy
appear optimum, very low doses comparable to annual background show dramatic hormetic
results. Azzam et al. [Rad. Res. 146, 369–371 (1996)] show that a dose of only 1 mGy to
mammalian cells in vitro reduced neoplastic transformations 3 or 4 fold below the spontaneous
rate. This may explain why people living in the seven states with the highest background have
about 15% lower cancer death rates than the average for the US (Frigerio et al. ANL/ES-26,
1973).
300
FIG. 1. Hypothetical dose-response curve for cancer induction by radiation

illustrating the potential for hormesis at doses below about 0.2 Gy.
Hormetic effects from radon progeny in the lung are suggested in the studies of Cohen (Health
Physics, Jan. 1995). The lung cancer mortality in US counties with the highest radon levels (>5.0
pCi/l) is about 40% lower than in US counties with the lowest radon levels (<0.5 pCi/l). This
suggests that the radiation eliminates cancers initiated by smoking.
Rebutal
Radiation hormesis is not a model—it is an effect. If a low dose is beneficial, it is hormesis. No

scientific definition is necessary. The scientifically respectable position is to report all data even
if they contradict current radiation protection dogma. [Radiation biologists have neglected
studies of biological effects at low doses (<0.2 Gy) where hormesis is most apt to be observed
because research money has been available to look for carcinogenic effects of radiation.]
Against the proposition: John E. Moulder
Opening Statement
The concept of ‗‗radiation hormesis‘‘ is certainly not ‗‗respectable‘‘ in health physics circles. It
has been condemned as unproven and lacking biological/biophysical plausibility, and criticized
because it would require a fundamental change in the basic radiation protection paradigm that
‗‗even the lowest radiation dose is harmful.‘‘ The concept is, of course, anathema to those who
claim that current radiation protection standards underestimate the risks of low dose exposure.
But none of these arguments provides a scientific basis for rejecting the concept. The lack of
hard evidence and established mechanisms does not make a concept unacceptable, it simply
makes it unproven; and the issue of whether a concept is ‗‗politically correct‘‘ should be
irrelevant to a scientific debate.
The only valid criterion for determining whether a concept is ‗‗scientifically respectable‘‘ is
whether the concept is sufficiently well-defined that it can be tested; that is, whether the concept
generates unambiguous hypotheses that are capable of being rejected. Here there is a problem
with ‗‗radiation
301
hormesis,‘‘ as the very concept is rather elusive. In a 1987 issue of Health Physics, Jerry Cohen1
defined ‗‗radiation hormesis‘‘ as the ‗‗process whereby low doses of [ionizing radiation]...could
result in stimulatory or beneficial effects.‘‘ However, two pages later Leonard Sagan2 defined it
as the presence of ‗‗effects unrelated to and unpredictable from the [effects of] high dose
exposure,‘‘ and still elsewhere Sagan3 argues that ‗‗radiation hormesis‘‘ could be equated with
adaptive response.4 To another strong proponent of the concept, T. D. Luckey, 5 ‗‗radiation
hormesis‘‘ is a specific type of nonmonotonic dose-response function that results in ‗‗beneficial‘‘
or ‗‗biopositive‘‘ effects at doses below 0.5 Gy.
These various statements of the concept are neither well-defined nor entirely compatible. What is
a ‗‗low dose?‘‘ What is ‗‗stimulatory‘‘ effect? What are ‗‗beneficial‘‘ or ‗‗biopositive‘‘ effects?
What are the criteria for establishing that effects are ‗‗unrelated‘‘ and/or ‗‗unpredictable?‘‘ In
fact, the concept of ‗‗radiation hormesis‘‘ is so vague that it is effectively impossible to
formulate unambiguous and testable hypotheses based on it. It is this vagueness which makes the
concept of ‗‗radiation hormesis‘‘ scientifically unacceptable.
Certainly there are circumstances where exposure to low doses of radiation confers benefits:
• Low doses (by radiation oncology standards) of total body irradiation are beneficial in bone
marrow transplantation.
• Some radiation-induced mutations can confer adaptive advantages on cells grown under
nonoptimal conditions.
• Doses of the order of 0.01–0.10 Gy can briefly confer resistance to subsequent higher doses of
radiation.
All of the above would appear to fit some of the definitions of radiation hormesis, but the
researchers who work in these areas see no need to invoke ‗‗radiation hormesis‘‘ to explain their
results.
Why is the concept of ‗‗radiation hormesis‘‘ so poorly defined? The answer is that proponents of
‗‗radiation hormesis‘‘ are largely reacting against proponents of the ‗‗linear no-threshold
hypothesis,‘‘ and the latter concept is itself essentially untestable at the doses that are of interest
in radiation protection. However, these two concepts are not alternatives, since low doses of
ionizing radiation could produce no detectable effects without implying hormesis. In fact, we
already know that sufficiently low doses of ionizing radiation cause no detectable effects. The
argument becomes whether these undetectable effects are beneficial or harmful—an argument
that harkens back to the question of ‗‗how many angels can dance on the head of a pin.‘‘
No model—whether it be ‗‗radiation hormesis‘‘ or the ‗‗linear no-threshold hypothesis‘‘—that is

based on the shape of dose-response curves for doses below the level where robust effects can be
detected should be considered scientifically respectable unless it is backed by well-understood
biological and biophysical mechanisms. The scientifically respectable position is to focus on
understanding the biological/biophysical basis for radiation effects; and that requires data, not
modeling or arguing about ill-defined concepts.
Rebuttal
While Dr. Cameron and I agree that the concept of ‗‗radiation hormesis‘‘ is not politically
respectable, we agree on little else. He does not share my concern that the concept of ‗‗radiation
hormesis‘‘ is poorly defined, yet he offers two additional definitions of the concept. First, he
would define hormesis in terms of ‗‗health benefits,‘‘ thereby restricting the scope to human (or
at least to plant and animal) studies.
302
Still later, his argument implies that the concept of radiation hormesis is equivalent to the claim
that ‗‗low‘‘ doses of ionizing radiation can protect mammals from naturally occurring cancer.
Neither of these definitions is fully compatible with those of Cohen, 1 Sagan,2,3 or Luckey.5 Thus
Dr. Cameron further reinforces my contention that the concept of radiation hormesis has not yet
been sufficiently well-defined that it can generate hypotheses that are both unambiguous and
rejectable.
The remainder of Dr. Cameron‘s argument is more an attack on blind adherence to ‗‗linear no-
threshold‘‘ models than it is an argument in favor of ‗‗radiation hormesis.‘‘ He argues that the
general applicability of ‗‗radiation hormesis‘‘ is proven by the observation that some data points
on some radiation carcinogenesis dose-response curves fall below the ‗‗zero dose‘‘ levels. This is
logically equivalent to arguing that the observation that some points fall above a ‗‗linear no-
threshold‘‘ line proves that the conventional ‗‗linear nothreshold‘‘ model underestimates low
dose risks. There is far more to risk assessment than arguing about the fit of small data subsets to
arbitrary models.6 It is time to put the modeling aside and focus on understanding the biophysical
and biological mechanisms that are responsible for radiation injuries. The modeling and the
theoretical arguments about modeling are starting to get in the way of the science.
REFERENCES
1. J. J. Cohen, ‗‗Conference on radiation hormesis: An overview,‘‘ Health Phys. 52, 519

(1987).
2. L. A. Sagan, ‗‗What is hormesis and why haven‘t we heard about it before?‘‘ Health
Phys. 52, 521 (1987).
3. L. A. Sagan, ‗‗On radiation, paradigms and hormesis,‘‘ Science 245, 574 (1989).
4. S. Wolff, ‗‗Is radiation all bad? The search for adaptation,‘‘ Radiat. Res. 131, 117–123
(1992).
5. T. D. Luckey, Radiation Hormesis (CRC, Boston, 1991).
6. K. R. Foster, L. S. Erdreich, and J. E. Moulder, ‗‗Weak electromagnetic fields and cancer
in the context of risk assessment,‘‘ Proc. IEEE 85, 731–746 (1997).
303
8.5. The use of low dose x-ray scanners for passenger screening at
public transportation terminals should require documentation of
the “informed consent” of passengers
Allen F. Hrejsa and Morris L. Bank
OVERVIEW
Government agencies are considering use of low-dose x-ray systems to scan all passengers (or
perhaps "suspicious" passengers) at transportation terminals. The agencies do not intend to seek
passenger consent because the screening dose is low. One can argue, however, that since
passengers do not benefit personally from screening, there is not a commensurate benefit to the
risk—and therefore passengers should consent before screening occurs. This issue is addressed in
Arguing for the proposition is Allen F. Hrejsa, Ph.D. Dr. Hrejsa received his Ph.D. in low energy
nuclear physics from the University of Notre Dame. He is the diagnostic medical physicist and
radiation safety officer at the Advocate Lutheran General Hospital as well as a consultant
medical physicist, primarily in the area of mammography. Dr. Hrejsa is certified by the American
Board of Radiology in diagnostic and therapeutic radiologic physics. He currently serves as
chairman of the governor's Radiation Protection Advisory Board for the state of Illinois. Dr.
Hrejsa has served on the AAPM Ethics Committee for many years and was chairman of AAPM
ethics committee from 1994 to 2000.
Arguing against the proposition is Morris L. Bank, Ph.D. Dr. Bank received a Ph.D. in Physics
from the University of Michigan and completed a postdoctoral fellowship in Radiological
Physics at the University of Wisconsin. Presently he is Associate Professor in the Department of
Radiation Oncology at Indiana University and Chief Physicist at the VA Medical Center in the
Department of Radiation Oncology. He is certified by the American Board of Medical Physics
and the American Board of Radiology.
FOR THE PROPOSITION: Allen Hrejsa, Ph.D.
Opening Statement
Currently there are two types of x-ray based body scanners. One type uses low energy
backscattered x-rays to produce a surface image of a person sans clothing. This machine can
image items concealed next to the skin.1 The second type is a transmission x-ray scanner which
can see objects that might be secreted in body cavities. I will address the backscatter devices
which deliver an effective dose of 0.03 µSv per anterior scan and between 0.01 and 0.02 µSv for
a posterior scan.2
The dose from the scanner is very low. The NCRP estimates that an individual would have to be
scanned 2500 times in one year to reach the administrative control limit of 0.25 mSv/year.2
304
Unfortunately the public has an inordinate fear of radiation stemming from information about
radiation in the public press and the entertainment media (e.g., Spiderman and the Incredible
Hulk). In addition, past mistakes by the government during the Atomic Bomb Testing and Civil
Defense era have left lingering fears and mistrust of assurances about the "harmlessness" of small
radiation doses.
Since many individuals would be exposed to this small radiation dose, body scanners have the
potential to measurably raise the collective dose to the population. We need to be forthright with
the public and inform them fully if x-ray scanners are going to be used for security purposes in
airports and other transportation venues. To this end, it seems reasonable to seek informed
consent from each individual who is to be scanned. If a signed informed consent is obtained, then
the individual would not be able to claim years later that he was unaware that he was being
exposed to x rays when he was scanned by the security system in the airport.
Another issue has been raised by the American Civil Liberties Union. This group opposes body
scans because "passengers expect privacy underneath their clothing and should not be required to
display highly personal details of their bodies ."3 In newspaper stories about the screening
trials run at the Orlando Airport, the process was characterized as a "peep show" or "virtual strip
search."4,5 Although privacy is a sensitive issue, an informed consent would describe the nature
of the image, how it was made, and any potential risks involved in its production. Manufacturers
are now going back to the drawing board and adjusting the software to insert "fig leaves" at
appropriate places in the images. Perhaps a concerned individual could be given the choice of an
x-ray image or a "pat down". Personally, I prefer the x-ray scan, as do several people I have
questioned (see also CBSNews.com July 17, 2003).6
Rebuttal
If low dose x-ray scanners are used for security purposes on a large scale, a signed consent would
provide a "paper trail" for future research into the effects of low doses of radiation to a large
population. Although record keeping would be a challenge, large-scale computing systems are
available which would allow collection of the data using "electronic" signatures on a
computerized informed consent form. At this stage in the research of low-dose radiation effects
in humans (less than 1 cGy per year), the consequences of low-dose radiation 20 years in the
future are unknown. So it is not possible to unequivocally state that the increased security
outweighs the radiation risk associated with low-dose x-ray scanners.
AGAINST THE PROPOSITION: Morris Bank, Ph.D.
Opening Statement
Government agencies and other institutions are proposing the use of whole body x-ray scanners
(WB scanners) for security purposes such as admission to public transportation and detection of
contraband in prisons. This use would involve repeated x-ray exposures of large numbers of
people with associated hazards. Proposed scanners use a scanning x-ray beam at 60–125 kVp to
produce images using backscattered x rays. The images show the skin surface of the person and
can identify weapons and contraband concealed under clothing.
There are two classes of WB scanners, General Purpose and Limited Use, with General Purpose
scanners proposed for security screening. The whole body effective dose of General Purpose
305
scanners is very low –0.1 µSv (0.01 mrem) according to the manufacturer and the National
Council on Radiation Protection and Measurements.2 A person could be scanned 2500 times
before reaching the recommended administrative control dose of 0.25 mSv (25 mrem) per year
[25% of max dose of 100 mrem]. Further, a person would have to experience 1000 scans to
achieve the annual Negligible Individual Dose [NID] of 0.1 mSv [10 mrem]. The scanners
conform to an ANSI standard for security screening7 of an effective dose below 0.1 µSv
(0.01 mrem) per scan. Such a low WB dose minimizes the radiation risk to persons undergoing
exposure. A traveler would require 50 scans per week to accumulate the maximum annual
effective administrative dose.
Passengers should be informed in writing of the use of an x-ray scanner, the exposure involved,
and the associated risks. However, informed consent and its documentation are not necessary.
Record keeping would be impractical considering the many locations of the scanners and the
many travelers screened. The benefit of mass scanning in terms of increased security greatly
outweighs the individual and collective radiation risk. This opinion is contingent upon instituting
additional measures such as shielding to limit radiation exposure to operators and bystanders.
Frequent inspection of units to ensure minimum dose levels, and periodic training for operators
should be included. Limited Use machines, which have higher effective whole body doses by a
factor of 10, are excluded from consideration in this proposal. For these units, record keeping and
protocols for limiting annual doses from repeated exposures would be necessary.
Rebuttal
I agree that the public must be knowledgeable and informed about whole body x-ray screening
procedures. A choice between a "pat down" and an x-ray procedure is a viable option. If an x-ray
scan is chosen, a description of the scanning procedure, the radiation dose and the associated
risks must be presented to all persons being scanned. A net benefit to society should result from
any x-ray procedure performed on humans. In this case the benefit is the safety of the passengers.
This is also the position of the Health Physics Society (HPS)8 concerning whole body x-ray
screening procedures.
Documentation of informed consent is not necessary, however, in order to inform the public of
the benefits and low risk of x-ray scans. In fact, screening procedures offer an opportunity to
educate the public about the benefits of radiation, and for medical/health physicists to be
involved in the discussion. The HPS has a stated position on whole body x-ray scanners, and the
American Association of Physicists in Medicine (AAPM) should also consider taking a position.
REFERENCES
1. W.J. Baukus, "X ray Imaging for On-The-Body Contraband Detection" Presented to 16th Annual
Security Technology Symposium & Exhibition, June 28, 2000, http://www.as-e.com.
2. NCRP Commentary No 16: Presidential Report on Radiation Protection Advice: Screening of
Humans for Security Purposes Using Ionizing Radiation Scanning Systems, December 15, 2003.
3. F. Reed, "Scanner virtually disrobes passenger," The Washington Times, May 22, 2003,
http://www.washingtontimes.com.
4. L. Ackerman, "Biometrics and Airport Security," Transportation Research Board (TRB) panel on
Personal Security, Washington D.C., January 13, 2003, http://www.privacyactivism.org/Item/64.
5. CNN "See-through scanner sets off alarms,"
http://www.cnn.com/2002/TRAVEL/NEWS/03/18/rec.airport.xray/?related.
6. L. Cowan, "New Airport X-Ray Too Revealing?" CBSNEWS.com, July 17, 2003.
306
7. ANSI (2002) "Radiation Safety for Personnel Security Screening Systems Using X-Rays"
ANSI/HPS N43.17-2002.
8. Health Physics Society, "Use of Ionizing Radiation for Security Screening Individuals,"
February, 2003.
307
8.6. The fetal dose limit for flight personnel should be 1 mSv over
the gestation period, rather than 5mSv as recommended by the
NCRP
Robert J. Barish and Richard L. Morin
OVERVIEW
An upper limit of 1 mSv fetal dose over the gestation period is used for employees of European
air carriers. This limit, endorsed by the ICRP and the 25 countries of the European Union, is
1/5th of the limit of 5 mSv over the gestation period used by US air carriers based on
recommendations of the NCRP. This discrepancy is the subject of this month's
Point/Counterpoint.
Arguing for the proposition is Robert J. Barish, Ph.D. Dr. Barish is a medical and health physicist
consultant in New York City. He received a B.S. in physics and a Master of Engineering in
radiological health from New York University and earned a Ph.D. in Medical Physics from the
University of London's Institute of Cancer Research. He was an Associate Professor of
Radiology at NYU Medical School, then chief radiotherapy physicist at the Cancer Institute of St.
Vincent Catholic Medical Centers of Brooklyn and Queens. His book, The Invisible Passenger:
Radiation Risks for People Who Fly, is used by airline crewmembers learning about in-flight
radiation. Dr. Barish is certified by the ABR, ABHP and ABMP and is a Fellow of the AAPM.
Arguing against the Proposition is Richard L. Morin, Ph.D. Dr. Morin is the Brooks-Hollern
Professor, Mayo Medical School. Dr. Morin received his Ph.D. from the University of Oklahoma
in Medical Physics. His dissertation concerned the use of Monte Carlo Simulation and Pattern
Recognition for artifact removal in Computed Tomography. He was Director of Physics in
Radiology at the University of Minnesota before joining Mayo Clinic in 1987. He is a Fellow of
the ACR and the AAPM and a Diplomate of the ABR. Dr. Morin is a former President and Chair
of the Board of the AAPM, and is currently a member of the ACR Board of Chancellors, chairing
the Commission on Medical Physics.
FOR THE PROPOSITION: Robert Barish, Ph.D.
Opening Statement
With regard to radiobiological effects, there is no test that would provide even the most careful
scientist with a means for determining whether an individual was exposed to 1 mSv or 5 mSv
distributed over a nine-month period. The issue we are debating is not one of demonstrable harm.
Rather, it is an examination of the logic used in establishing standards for radiation protection at
low levels, in particular for the radiation received in high-altitude air transport. Unlike other
occupations, where actual exposures are often considerably below administrative limits, the
cosmic radiation dose received by crewmembers in an airliner is an inevitable consequence of
308
their work environment. For pregnant crewmembers on high-altitude routes, the dose will exceed
1 mSv over the gestation period of nine months, assuming normal work schedules.1
At these levels there are no deterministic (non-stochastic) risks. At such low doses, however, the
linear no-threshold (LNT) model of radiation risk predicts a greater-than-zero chance of a
malignant transformation that might cause a cancer many years after exposure. The LNT model
has recently been upheld by the National Council on Radiation Protection and Measurements
(NCRP) in Report No. 136.2 The risk of a radiation induced childhood cancer for the offspring of
an exposed crewmember cannot be ruled out.
In Report No. 39 in 1971, the NCRP established dose limits for individuals exposed to ionizing
radiation.3 The dose limit for members of the public was set at 5 mSv per year. In that report the
Council specifically stated that: "the maximum permissible dose equivalent to the fetus from
occupational exposure should not exceed 0.5 rem." Thus the NCRP recommended an identical
exposure limit for the embryos and fetuses of occupationally exposed women as for members of
the general public, specifically pointing out the fact that "embryos, fetuses and young children"
would be included in that latter category.
When NCRP Report No. 91 appeared sixteen years later,4 the maximum permissible dose to
members of the public was lowered to 1 mSv per year. But the allowed dose to the embryo/fetus
of occupationally exposed women was not reduced to that lower level. It was kept at 5 mSv.
Thus the categorization of embryos and fetuses of occupationally exposed women was changed,
from essentially being members of the public to an in-between status with an exposure limit five
times greater than members of the public and ten times lower than an occupationally-exposed
pregnant woman. Identical limits appear in a subsequent NCRP Report No. 116.5
About the time that the NCRP published Report No. 91, the International Commission on
Radiological Protection (ICRP) produced Publication 60 in which a fetal dose limit of 1 mSv was
established,6 a value essentially matching the public dose limit. All twenty-five nations in the
European Union accept this value in regulating exposures of pregnant aircrew. The United States
Federal Aviation Administration uses it as well in its advisory documents.1
It follows from the assumptions that the stochastic risk of radiation exposure at low levels is not
zero, that the LNT model is applicable and that the NCRP recommendations permit a five-times-
greater risk of childhood leukemia or other malignancy for the child of an air crewmember
working for a US air carrier than for a similarly employed woman working for a European
airline. I don't think this is appropriate.
Rebuttal
Dr. Morin and I both agree that, at the levels of exposure under discussion, there can be no
scientific test of demonstrable harm. He argues that a change in permissible exposure from 5 mSv
to 1 mSv for pregnant flight crewmembers would "have tremendous societal and economic
impact without justification." But he does not back up this assertion with any statistic on the
number of flight attendants and pilots who actually become pregnant each year, nor with the
actual economic impact on an airline if they were assigned ground-based activities as an
alternative to their in-flight duties. He ignores the fact that all European carriers presently adhere
to the lower dose limit yet remain economically viable.
309
With respect to frequent-flying passengers, it is hard to imagine that any woman would make
twenty transcontinental trips or seven intercontinental journeys during pregnancy as leisure or
vacation activity. A flight schedule like that would certainly be a consequence of their
employment. These women, like the aircrew they travel with, are also occupationally exposed
individuals. My arguments regarding aircrew dose equally apply to them!
It is a matter of law that any restriction of radiation during pregnancy requires a declaration by
the exposed woman that she wishes to have lower dose limits applied. A woman who does not
wish to have these restrictions may simply opt out of the required declaration. So it would not be
the airlines that would deny a frequent-flying passenger a ticket. Rather, it would be a matter for
employers to put into place for these women (and indeed for their male colleagues) the same type
of radiation training required for traditional radiation workers in other settings to help them make
informed decisions about their occupational exposures. Business frequent flyers, exposed to
cosmic radiation in airliners, should be educated in the same manner as workers who are exposed
to ionizing radiation in other occupations.
AGAINST THE PROPOSITION: Richard Morin, Ph.D.
Opening Statement
This proposition suggests that the radiation exposure limit for pregnant airline flight personnel
should be lowered. Why? Are there scientific reports of radiation induced anomalies, morbidities,
or mortalities for the children of flight personnel? Are there unscientific reports of such
occurrences? I believe the posing of such a proposition is simply a manifestation of an overall
"lower is better" philosophy regarding radiation exposure. For the sake of discussion, let's ignore
the fact that there are no data to suggest current activities require a change in the limit and
examine the consequences of such a change.
For an intercontinental trans-polar roundtrip flight from New York to Tokyo, an exposure of
about 0.15 mSv is expected7 from calculations using FAA software8 that accounts for altitude,
latitude, and typical flight time. With the lowered limit, a pregnant employee would be limited to
about 7 such trips and no other flights during her pregnancy. Alternatively, about 17
transcontinental New York to Seattle trips7 with no other trips would be allowed (0.06 mSv/trip)
during pregnancy. Restricting flight personnel in such a manner would imply that they could only
work about 30 days during their nine-month pregnancies. The consequences would have a
tremendous societal and economic impact without justification. However, this is just the tip of
the iceberg. If a lower limit is appropriate for safety reasons for flight personnel, then it must also
apply to "precious metal" frequently-flying passengers (probably including some reading this
column) who easily accrue the same amount of flight time. Who would enforce this restriction? Is
it even remotely possible that an airline would deny a passenger a ticket because she had 20
transcontinental trips by day 90 of her pregnancy? Who would then tell her that she couldn't fly
for six months?
A "lower is better" philosophy is not supported by science or common sense. It also reflects
contorted thinking that perpetuates public misunderstanding and fear of radiation and its
biological effects. Unfortunately, when the risks and benefits of ionizing radiation are raised in a
public setting, most persons owe their knowledge of this topic to Homer SimpsonTM!
310
I argue against this proposition because it has no scientific basis, the consequences to society are
enormous, and most importantly, it perpetuates an untruth to the public for no reason at all.
Rebuttal
I compliment my colleague on a scholarly position statement. Academic Theory (somewhat like

Legal Theory) is always interesting. However, the topic we are debating is whether or not the
current exposure limit should be lowered. Is there evidence for an increased incidence of harm
for airline flight crews compared with airline workers who don't fly? I can find none. My
colleague has invoked the Linear Non-Threshold Hypothesis (LNTH) for the Biological Effects
of Ionizing Radiation. This hypothesis is not without controversy.9,10,11,12 Not all experts agree
with its use at the level of radiation dose experienced by air crews. The fact that the European
Union has adopted an ICRP recommendation based on the LNTH doesn't verify the LNTH or
establish this course of action as correct. The LNTH is, after all, a hypothesis. Perhaps the
European Union has it wrong. It is time for the scientific community to meet its responsibility to
demand data in support of regulations. Health care is pushed to provide "evidence-based
medicine." We should join together and demand "evidence-based regulations."
I recognize that change is difficult (we still use Roman Numerals to designate Super Bowls). It is
time, however, for the scientific community to state unequivocally that the emperor is without
clothes. I return to my original statement regarding this proposition. The proposition suggests
that the radiation exposure limit for pregnant airline flight personnel should be lowered. Why? To
follow this course regardless of scientific knowledge, acumen, or insight, is illogical and
potentially of greater risk. As a practical matter, if there are no proven untoward effects, why
change and incur the problems and downside risks inherent in such an effort.
REFERENCES
1. W. Friedberg and K. Copeland, "What aircrews should know about their occupational
exposure to ionizing radiation," Available at: http://www.cami.jccbi.gov/AAM-
600/Radiation/trainingquestions.htm.
2. NCRP Report No. 136: Evaluation of the linear-nonthreshold dose-response model for
ionizing radiation (National Council on Radiation Protection and Measurements, 2001).
3. NCRP Report No. 39: Basic radiation protection criteria (National Council on Radiation
Protection and Measurements, 1971).
4. NCRP Report No. 91: Recommendations on limits for exposure to ionizing radiation
(National Council on Radiation Protection and Measurements, 1987).
5. NCRP Report No. 116: Limitation of exposure to ionizing radiation (National Council on
Radiation Protection and Measurements, 1993).
6. International Commission on Radiological Protection, 1990 Recommendations of the
ICRP, ICRP publication 60, In Annals of the ICRP 21 (1-3). (Pergamon Press, 1991).
7. R. J. Barish, "Radiation risk from airline travel," J. Am. Coll. Radiol. 1, 784–785 (2004).
8. Federal Aviation Administration (U.S.). Radiobiological Team Website. Computer
program CARI-6. http://jag.cami.jccbi.gov./cariprofile.asp.
9. G. M. Matanoski, A. Sternberg, and E. A. Elliott, "Does radiation exposure produce a
protective effect among radiologists?" Health Phys. 52, 637–643 (1987).
10. R.J. Barish, "The Invisible Passenger: Radiation Risks for People Who Fly," (Advanced
Medical Publishing, Madison, 1996).
11. "Health effects of low-level radiation: Position statement 41," (American Nuclear
Society, 2001), http://www.ans.org/pi/ps/
311
12. R.L. Dixon, J.E. Gray, B.R. Archer, and D.J. Simpkin, "Modern radiation protection
standards: Their evolution from science to philosophy," Radiat. Prot. Dosim. 113: (in press).
312
8.7. A pregnant resident physician should be excused from training

rotations such as angiography and nuclear medicine because of the
potential exposure of the fetus
Edward L. Nickoloff and Libby Brateman
Reproduced from Medical Physics, Vol. 26, No. 12, pp. 2517-2519, December, 1999
OVERVIEW
It has been reasonably well documented that a pregnant resident physician can assume radiology
rotations, including higher-exposure rotations such as angiography and nuclear medicine,
without exposing the fetus to radiation levels that exceed national and international guidelines.
Hence, many medical physicists support the contention that rotations should not be altered
because a resident is pregnant. On the other hand, many if not most physicists subscribe to the
ALARA (as low as reasonably achievable) principle, especially in cases of fetal exposure where
increased radiation susceptibility is combined with an inability to decide for oneself. In addition,
altered rotations usually can be accommodated by swapping rotations with other residents, with
the pregnant resident taking high exposure rotations after delivery of the child. Policies on this
issue vary among institutions, possibly because medical physicists have not come to closure on
the issue. This issue of Point/Counterpoint is directed toward that objective.
Arguing for the Proposition is Edward L. Nickoloff, D.Sc. Dr. Nickoloff received his Doctor of
Science Degree with Distinction from The Johns Hopkins University in 1977 after which he
served the Department of Radiology at Johns Hopkins as Acting Director of Physics and
Engineering. Currently, Dr. Nickoloff is Professor of Clinical Radiology and Chief Hospital
Physicist at Columbia University P&S and the New York-Presbyterian Hospital (Columbia-
Presbyterian Center). Dr. Nickoloff is board certified by the ABR, ABMP, and the ABHP, and a
Fellow of AAPM, ACMP, and ACR. His research interests include image quality assessment,
quality control, radiation dosimetry and radiation shielding, physics instrumentation and
technical aspects of CT/mammography/ digital systems.
Arguing against the Proposition is Libby F. Brateman, Ph.D. Dr. Brateman has been a medical
physicist for 25 years in private, government, and university-related hospitals and at the former
BRH. As a member of the AAPM Radiation Protection Committee, she has championed
reasonable radiation protection regulations for x-ray workers at both state and national levels.
She is Associate Professor of Radiology at the University of Florida College of Medicine. A
member of the ACR Committee on Mammography Physics, she is a five-year breast cancer
survivor. As the first adult ventilator survivor in her Bone Marrow Transplant Unit, she
understands the dangers of emphasizing statistics in predicting health outcomes for individuals.
FOR THE PROPOSITION: Edward L. Nickoloff, D.Sc.
Opening Statement
Radiation protection criteria for the human fetus are based upon risk estimates derived from
limited data on atomic bomb survivors, a few retrospective clinical surveys and animal studies.
313
These risk estimates contain considerable uncertainty, especially at low dose levels.
Nevertheless, the literature does identify a number of risks from fetal exposure to ionizing
radiation, such as: leukemia, other cancers, growth retardation, microcephaly, diminished
intelligence, mental retardation, genetic mutations, and other effects.
In-utero irradiation of a human fetus tends to affect the Central Nervous System (CNS) rather
than to cause the organ and limb anomalies seen in animal experiments. Moreover, extrapolation
of animal data to predict human fetal response to radiation has many caveats.
Recent radiation biology studies suggest that cancer can arise from a single cell that has
undergone mutations in oncogenes and/or deletion of suppressor genes. Furthermore, fetal cells
are more sensitive to radiation damage than other cells because of their increased rate of mitosis,
especially during organogenesis. The cancer risk from in-utero radiation exposure has been
estimated in NCRP 115 and BEIR V as about 0.025% per cGy. Even at the regulatory limit for
the fetus of 0.50 mSv per month, the lifetime cancer risk from fetal radiation exposure would be
about 0.01%. Because it is a stochastic process, however, radiation induced cancer could occur
even at the lowest exposure levels (no threshold model of radiation injury).
Other potential risks to the fetus from in-utero irradiation include severe mental retardation with
an estimated risk of 10%–40% per Gy, diminished IQ estimated at 225 points per Gy, and
genetic induced defects estimated at 1% per Gy. Other effects to the CNS from fetal irradiation
may be difficult to assess, such as: functional or behavioral defects, emotionality, impaired
nervous reflexes, hyperactivity, and various learning deficits.
Unlike patients who receive benefits from medical procedures like angiography that involve
significant levels of radiation, the fetus of the pregnant physician is exposed to risk without any
benefit. In particular, interventional angiography and cardiac procedures require considerable
fluoroscopy and deliver significant radiation dose to patients, and relatively high cumulative
scattered levels to staff. Radiation levels involved with these procedures are so large that patient
radiation injuries have been reported. Subjecting pregnant physicians to the mental stress and
guilt from potential adverse effects to their babies is unconscionable, can easily be avoided, and
is not consistent with radiation protection policy.
It is prudent to be conservative and to employ the principles of ALARA. In many facilities, it is

feasible to temporarily re-assign pregnant resident physicians to other duties in order to limit the
radiation exposure to the fetus, allowing them to complete their high-exposure training rotations
at a later date. For these reasons, pregnant resident physicians should be temporarily excused
from clinical training rotations that could potentially expose the fetus to significant levels of
ionizing radiation.
Rebuttal
Dr. Brateman has raised a number of different issues against the proposition. The regulatory limit
of 0.5 mSv per month to the fetus of a pregnant radiation worker is not a magic number below
which no detrimental effects could occur. Many of the adverse effects to the fetus are stochastic
processes by which a single radiation-damaged cell could result in cancer or other problems, i.e.,
there is no threshold level. Moreover, the radiation risk is significantly greater to the fetus than to
adults. Regulatory limits for the maximum permissible dose for occupational exposure are
subject to change and have been reduced over the years from 100 mSv per day in 1902 to the
current value of 0.20 mSv per day due to a better understanding of radiation biology.
314
Indeed, the risks to the fetus from in-utero irradiation may be relatively low. By analogy, the risk
of an individual getting struck by lightning is also relatively low. Nevertheless, lightning kills
60–80 persons in the U.S. each year. It is still prudent to use caution both for persons venturing
into a thunderstorm and for pregnant residents working in high radiation areas—regardless of the
relatively low risk.
It is commendable that Dr. Brateman‘s clinical facilities utilize radiation protection practices that
maintain very low exposures to their staff. Nevertheless, some procedures in radiology can
expose staff to high levels of radiation. A single interventional angiography or cardiac procedure
may require 30–90 min of fluoroscopy. Moreover, some nuclear medicine facilities are involved
with therapeutic radioisotope procedures (labeled antibodies and thyroid ablations) and with
cyclotron ‗‗hot labs.‘‘
The social, economic, and political ramifications associated with this issue are indeed complex.
The 1991 Americans with Disabilities Act provides protection from discrimination against
pregnant workers; and the 1993 Family and Medical Leave Act has provisions for new mothers
and fathers. I would also hope that a number of highly publicized lawsuits have sensitized
employers to sex discrimination. Dr. Brateman is a highly qualified and competent medical
physicist. Although our positions on this topic are diametrically opposed, I do understand and
respect her position.
AGAINST THE PROPOSITION: Libby Brateman, Ph.D.
Opening Statement
Excusing pregnant residents from selected rotations is not a recommended policy, because it is
not justifiable and potentially disruptive. Data from our institution show physicians performing
angiography receive under-apron dose equivalents which are less than 0.5 mSv a month in
almost all cases. (The few exceptions have been for fellows who performed interventional
procedures for more than 40 h a week.) Attenuated conceptus doses are lower, which could be
further reduced with maternity aprons or mobile shielding. Our nuclear medicine physicians
never receive measurable monthly doses. Therefore, nearly all physicians receive doses within
legal limits for pregnant individuals, without modification to existing practices.
The first implication of excusing pregnant residents from certain assignments is that these
assignments pose greater risk than others. Radiation effects have not been demonstrated for
doses as low as—or even close to—the occupational limits for individuals with declared
pregnancy. Policies and regulations are based on the recommendations of international experts.
For a facility to ignore them is difficult to justify from a legal standpoint. If expected
occupational radiation doses for certain rotations are considered too risky, even though within
guidelines and regulatory limits, what criterion is to be used as the basis for policy?
A second implication of excusing pregnant residents is that these rotations would also be
hazardous for potentially pregnant residents. It is not possible to know immediately that
conception has occurred. What inference is to be made by the resident already in such a rotation
when she learns of her pregnancy? Is there ever a safe time for potentially pregnant residents?
315
Residencies are organized for residents to complete their training in a general order, with more
demanding modalities typically scheduled later in the program. Requested schedule changes may
cause negative feelings toward the pregnant resident, which may be reflected in evaluations that
affect her career and which put her in jeopardy for having done so— particularly in comparison
with a pregnant resident who does not request to be excused.
Excusing pregnant residents also implies that other pregnant personnel (e.g., physicians,
technologists, nurses) should be excused. Unlike residents, these individuals may be specialized
so that no reasonable substitute assignment exists for them. Difficult to manage, this situation
may lead to unlawful discrimination against females because of potential pregnancy.
Regulations are set to reasonable limits. Facilities must ensure that these limits are followed, and
they must not discriminate unlawfully. Personnel policies which are justifiable and universally
applicable provide the best protection.
Rebuttal
Risk estimates for stochastic effects for pregnant worker doses contain considerable uncertainty
and include ‗‗no effect‘‘ as a possibility. A prudent assumption is that the radiation-related
cancer risk is real, even if overestimated. Therefore, following ALARA principles, NCRP Report
116 suggests using a linear dose-response model without threshold for the purpose of radiation
protection, with a limit of 0.5 mSv/month to the embryo/fetus. Deterministic effects such as
neurological impairment in offspring have not been seen for doses below 0.1 Gy, doses 20 times
higher than the limit.
Radiation protection regulations are based on recommendations of the NCRP, which include the
concept of justifying and limiting radiation exposure: That is, ‗‗the need to apply individual dose
limits to ensure that the procedure of justification and ALARA do not result in individuals or
groups of individuals exceeding levels of acceptable risk.‘‘ NCRP Report 115 states that, even
for the highly unlikely maximum dose to the woman (50 mSv), the risk is ‗‗small compared to
other risks to the fetus,‘‘ e.g., 4%–6% of live births with congenital abnormalities.
Cancer is common, with an approximate 41% incidence, which varies by type and incidence
among states. For example, lifetime breast cancer incidence (age-adjusted) varies among states
from 7.15% to 9.37%. This variability is 177 times greater than the 0.0125% lifetime risk for
radiogenic cancers at the maximum allowable dose to the embryo-fetus (0.025%/cSv x 0.5
mSv/month x 10 months).
Every occupation has risks, and everything we do has an associated risk. The probabilities
associated with these risks are largely unknown and depend on innumerable factors, many of
which are controlled by our choices in living our lives. Some choices are based on reason, with
comparisons of risks and benefits (or risks and risks), and some are not. Some of us are more
cautious, and what is frightening to some is fulfilling for others. No reasonable person
intentionally harms him/herself, family or others. No employer wants to subject employees,
pregnant or not, to mental stress and guilt associated with radiation exposure to themselves or
their offspring; therefore, appropriate regulations are the standard. As medical physicists, our
socially-responsible task is to educate residents, so that they are able to be assured in performing
their jobs as radiologists while maintaining ALARA. They will then become better physicians
and employers of others, including pregnant staff.
316
8.8. An occupancy factor of unity should always be used for waiting

rooms and other highly-occupied public areas
Douglas R. Shearer and Michael Yester

Reproduced from Medical Physics, Vol. 27, No. 9, pp. 2005-2007, September, 2000
OVERVIEW
The linear no-threshold model of radiation injury suggests that a small radiation dose delivered
to a large group has the same effect on a population as a large dose delivered to a small group,
provided that the person-sievert product is the same to the two groups. According to radiation
advisory agencies (e.g., the NCRP), an occupancy factor of one should be used for shielding
computations if a single member of the public continuously occupies an adjacent room.
However, if the room is occupied by several people over the same period, but no one individual
for the entire period, an occupancy factor of less than one may be used. One could argue that this
procedure violates the concept of the linear no-threshold model. In fact, one might suggest that
an occupancy factor greater than one should be used when the average number of persons
occupying the room exceeds unity. This issue is the topic of this month‘s Point/Counterpoint
series.
Arguing for the Proposition is Douglas R. Shearer, Ph.D. Dr. Shearer is an Associate Professor at
Brown University and teaches courses in Medical Imaging and Health Physics. He is also the
Director of Medical Physics at Rhode Island Hospital. Dr. Shearer is a Fellow in the American
College of Radiology. Dr. Shearer obtained an Honors Degree in Natural Philosophy (Physics) at
Glasgow University in Scotland and did postgraduate work in Radiation Physics at the
University of London where he received a M.Sc. degree with distinction and a Ph.D. He is the
author of approximately 100 scientific papers, book chapters and abstracts as well as an article
on single-handed ocean sailing.
Arguing against the Proposition is Michael Yester, Ph.D. Dr. Yester is a Professor in the
Department of Radiology at the University of Alabama at Birmingham. Dr. Yester is currently
Chair of the Nuclear Medicine Committee of the AAPM and serves on the Education Committee
and Summer School Subcommittee. Dr. Yester has been active in the Southeastern Chapter of
the AAPM and recently served on the AAPM Board of Directors as a chapter representative. Dr.
Yester is certified by the American Board of Radiology and is a Fellow of the American College
of Radiology and the AAPM.
FOR THE PROPOSITION: Douglas R. Shearer, Ph.D.
Opening Statement
The linear no-threshold (LNT) dose hypothesis when applied to stochastic radiation effects
implies three things. Firstly, there is no dose of radiation, however small, which does not carry a
risk. Secondly, that risk is linear with dose. Thirdly, the risk is independent of dose rate; thus 30
mSv to one person carries the same risk as 1 mSv to 30 persons. In fact, this concept of
317
collective dose is explicitly stated in NCRP Report 121. ‗‗The effect or risk of a given dose is
identical whether the collective dose is administered to a single individual or distributed over a
population of individuals.‘‘ However, in the application of occupancy factors for radiation
shielding design, we depart from the LNT concept.
Occupancy factors are defined rather vaguely in NCRP 49 as ‗‗The factor by which the workload
should be multiplied to correct for the degree of occupancy of the area in question while the
source is ‗ON‘.‘‘ More precisely, NCRP 127 defines the occupancy factor as the fraction of time
that a space will be occupied by any single individual.
The use of this factor can be illustrated by application to a patient waiting room adjacent to a
diagnostic or therapeutic x-ray installation. In NCRP 49, waiting rooms are defined as having
occasional occupancy and an occupancy factor of 1/16 is suggested. For a 40 h week, this
implies that one single person is present in that room for ~2.5 h. This seems rather high and
occupancy factors of ~1/40 have been suggested, i.e., 1 h/week. Assuming the permissible
exposure to a member of the public is 1 mSv/yr, a dose equivalent to the waiting room space of
16 or 40 mSv/yr would be possible.
However, if we are true to the ideas of LNT, we should consider that for a typical workweek,
there is probably at least one person (not the same person) in the waiting room and maybe many
more, depending on the capacity of the waiting room—perhaps as many as 10. Thus, the
collective dose could lie between 16 and 400 person mSv/yr.
The occupancy factor should then be a function of waiting room capacity and could vary
between 1 and 10 or more. This means that shielding requirements would be increased by a
factor of 16–400 (a few tenth value layers) to keep the collective dose to the recommended 1
mSv/yr. This would increase the cost of shielding by a significant amount but, according to LNT,
there is no alternative.
By assuming the currently accepted fractional occupancy factors, we have implicitly discarded
the LNT hypothesis.
Rebuttal
I have little argument with any of Dr. Yester‘s points. He is stating the practical and common
sense point of view. However, in the first sentence, he states that low levels of radiation are
‗‗safe.‘‘ The LNT hypothesis states exactly the opposite, i.e., there is no ‗‗safe‘‘ dose of
radiation. There is always some risk.
Based on the LNT and the collective dose concept, we have assigned thousands of cancer deaths
to the Chernobyl radioactive releases. The EPA currently informs the public that 14 000 lung
cancer deaths/yr (actually somewhere between 7000 and 30 000) are due to the inhalation of
small quantities of radon. In fact, by combining a mathematically derived and almost negligible
risk with a large population of susceptible individuals, almost any number of deaths can be
calculated.
It is not a tenable scientific position to apply LNT and collective dose concepts to some scenarios
and tacitly discard them for convenience in others.
318
The scientific community must adopt a consistent stance to the evaluation of risk from small
doses of radiation.
AGAINST THE PROPOSITION: Michael Yester, Ph.D.
Opening Statement
Shielding calculations are a common source of conflict between prudent safety and recognition
that low levels of radiation are ‗‗safe.‘‘ In this conflict, the magnitude of ‗‗low‘‘ is a major
subject of debate. Current guidelines stipulate that the general public should not receive more
than 1 mSv equivalent dose in a year over and above background excluding medical procedures.
(The typical accepted average background equivalent dose is 1 mSv excluding radon exposure.)
In any event, 1 mSv/yr (0.02 mSv/week) represents the target of calculations for shielding
calculations. It has been recognized that the limit can be modified by the use of an occupancy
factor. Although NCRP guidelines for shielding currently in use (NCRP 49) employ strange
combinations of groups and classifications for occupancy factors, these factors represent current
practice. The current guideline for a waiting room is an occupancy factor of 1/16 (occasional
use).
In this Point/Counterpoint, the linear no-threshold (LNT) model and the concept of collective
dose are under scrutiny. Although the use of the LNT model is controversial, it is widely
recognized as being conservative. There is significant evidence for other than a linear no-
threshold model at low levels of exposure, but the LNT model is the model to be used for
radiation protection purposes. If one is making use of the collective dose concept, NCRP 121
states that it should be used with care at low levels of exposure, and the population distribution
should be taken into account. In the case being considered, a general waiting room or similar
highly-occupied area will consist of a significant transient and variable population. Moreover,
the levels of radiation are quite small, so the application of the collective concept is questionable.
In the absence of a good model, it is useful to consider the exposure to an individual for the
calculations. Let us consider, then, the shielding from the perspective of an individual.
Realistically the time spent by a single individual in a waiting room is short, especially as a
typical hospital stay is quite brief these days. In addition individuals would tend to move around
within the area. An occupancy factor of 1/16 would represent 2.5 h over a given week, and actual
presence in such an area for multiple periods during the year by a given individual would be
quite rare. Furthermore, radiation barrier calculations are performed for a point at a distance of
30 cm from the wall, which is again a worst case scenario.
Since there is considerable conservatism already included in the calculations and exposure limits,
reasonable occupancy factors should be allowed. Such considerations provide for exposures well
within the standard guidelines and in fact contribute minimally to an individual‘s annual
exposure.
Rebuttal
Dr. Shearer concludes that fractional occupancy factors violate the LNT hypothesis. It is also
admitted that following this principle within the current context of debate would increase the cost
of shielding significantly. If this is the case, does not this imply that something is wrong when
such small exposures are in question? In NCRP 121, it is explicitly stated that the collective dose
319
concept should be applied with appropriate caution when the population characteristics are
poorly defined, or have a high degree of uncertainty, or are subject to significant temporal
fluctuations. It was stated that the range of exposure in a waiting room could vary by a factor of
10 or more which would make the uncertainties in the dose quite variable and unreliable.
Temporal fluctuations will be significant in common waiting rooms. Thus, it is reasonable to
state that the collective dose concept is not applicable in this situation. [It is interesting to note
that in 1984 the estimated collective dose equivalent for air travel in the U.S. was 2500 person-
Sv (NCRP 93).]
In a final analysis, consider actual shielding practice. The standard material for shielding
construction in diagnostic radiology areas is a thickness of 1.5 mm of lead. For a waiting area,
the adjoining wall should be a secondary barrier so that only scatter is important. As such, the
common thickness affords a significant degree of protection at a modest cost, and the actual
exposures in a waiting room are at background level. In my opinion there is little need to use a
greater thickness of lead in such imaging situations. However, if the use of reasonable occupancy
factors is needed to keep the shielding to a reasonable level, it should be considered rational to
do so.
320
8.9. States should develop regulations to require monitoring of

radiation doses during interventional fluoroscopy
Louis K. Wagner and Raymond L. Tanner

OVERVIEW
Several cases of radiation burns have resulted from excessive use of radiation during
interventional fluoroscopic procedures. These cases suggest that some physicians are unaware of
the amount of radiation being used. Monitoring of patient exposures during the procedures could
correct this shortcoming. It is unlikely that exposure monitoring would be implemented
voluntarily. Hence, regulatory action would be necessary. Whether such action is desirable is the
subject of this Point/Counterpoint.
Arguing for the Proposition is Louis K. Wagner, Ph.D., Professor of Radiology at the University
of Texas—Houston Medical School. Dr. Wagner is a Fellow of the ACR and the AAPM and a
member of the NCRP. He has served on many national and international committees, including
advisor to the NRC on the medical uses of isotopes. Dr. Wagner has published extensively on
radiation effects from diagnostic and interventional radiations. Publications include his books:
Exposure of the Pregnant Patient to Diagnostic Radiations and Minimizing Risks from
Fluoroscopic X Rays. Most recently he served as advisor and author on "Skin injuries from
fluoroscopically guided procedures"—Parts 1 and 2 that appeared in the July 2001 issue of the
American Journal of Roentgenology.
Arguing against the Proposition is Raymond L. Tanner, Ph.D. Dr. Tanner is Professor Emeritus
of Radiology at the University of Tennessee, Memphis. He is a Fellow of the AAPM and the
ACR, served as President of the AAPM when the Centers for Radiological Physics were
established, was the first physicist appointed to the ACR Board of Chancellors, and helped found
its Mammography Accreditation Program. His 40 year career encompasses college physics
instruction, resident and graduate student teaching, provision of radiology planning expertise and
consulting in radiation protection and quality control. He served on numerous local, state,
regional, and national committees and commissions and examined for the ABR for over 25 years.
Directly relevant to this proposition was his six year representation to the CRCPD on behalf of
the ACR.
FOR THE PROPOSITION: Louis K. Wagner, Ph.D.
Opening Statement
In 1996, over 700 000 fluoroscopically guided interventional procedures were performed in the
United States; more than 1 000 000 are expected this year. The value of these procedures in
saving lives is well established.
321
One of several risks associated with these procedures is radiation. Doses have been sufficient in
some patients to induce very painful skin injuries that have lasted for extended periods, from
months to years.
About 100 injuries can be documented, but there is a disturbing lack of knowledge as to how
prevalent they actually are. The most unsettling fact is that these injuries have been frequently
misdiagnosed as contact dermatitis, chemical burn, electrical burn, insect bite, etc. This is due in
large measure to the delay between irradiation and the appearance of a lesion. In some severe
cases, the progression of the injury was so baffling to physicians that a tenacious investigation
was launched. Sometimes years passed before the correct diagnosis was reached. Recently, six
cases at a single facility were identified.
Monitoring of radiation skin dose has not been required for fluoroscopy for at least two reasons.
In the past, doses from diagnostic procedures have not been sufficiently high to cause injuries.
This has led to a false sense of security that fluoroscopy is safe. The second is that technology to
monitor skin dose has not been available. Neither of these factors applies today to interventional
work.
Several facts make skin dose measurements necessary for some fluoroscopic procedures today:
Doses to skin can be sufficiently high to induce injury,
Skin dose depends on the operator's knowledge of how to apply fluoroscopy (and
fluorography),
Commercial fluoroscopic equipment varies tremendously in radiation output and radiation

abatement technology,
Physicians are unaware of how much radiation they deliver to patients,
Many, if not most, physicians are unaware that fluoroscopy can injure patients.
A regulation that requires interventionalists to estimate skin dose specific to their procedures and
equipment is necessary for proper patient care. Clearly this knowledge alone is not sufficient. A
quality control review of doses, with establishment of action thresholds, is necessary to prevent
excessive doses and injuries to patients. The FDA issued an advisory in 1994, with revisions in
1995, recommending dose monitoring. Few have heeded this advice. Since then, the number of
reported injuries has risen dramatically. The time has come to settle the issue. Since voluntary
compliance doesn't work, mandating such measurements through regulation is necessary.
Rebuttal
I agree with Dr. Tanner that education, training, quality control, appropriate equipment, and well-
designed facilities are the primary elements for fluoroscopic safety. In their 1994 warning, the
FDA cited many of these elements as essential. Little progress has been made since then. Most
interventional physicians have virtually no training in safe fluoroscopic practices, and they have
little appreciation for the amount of radiation they apply to a patient's skin. How can they know
when many machines have no dose-monitoring equipment? Even if available, many physicians
do not know how to interpret the dose-area-product readout. Consequently, the readout is
ignored.
322
All too often, equipment is inappropriately designed for interventional use. Many machines have
too little filtration and no variable-pulsed fluoroscopy. Even when available, physicians do not
take advantage of dose-saving measures. Some machines have ill-designed features that defeat
dose-saving tools. Patients deserve better than this.
Quality control and administrative rules do not solve the problem either. I applaud facilities that
have good quality control programs and require credentialing and training of fluoroscopists. Too
few facilities do this.
Therefore, based on a proven lack of success of Dr. Tanner's options (a)–(d), option (e),
government regulation, must be implemented. While I agree that a federal agency should oversee
such a requirement, no federal agency has the authority to do so. Only state agencies can provide
this regulation.
I mostly agree with Dr. Tanner's other comments about the folly of the regulatory process in the
United States. Regulators and physicists benefit by over regulation because it creates job
security. To support regulation for this reason is appalling and unethical—but it occurs. On the
other hand, I cannot support an argument to avoid essential regulation out of fear of over
regulation.
The problem with regulation is enforcement, not the rules themselves. Enforcement policies that
obsess on ferreting out "violations" while ignoring the overall quality of a program are the
primary villain. This binary approach to total compliance or violation is silly. A system that
grades the quality of a program with a pass/fail result, together with a list of recommendations
and/or commendations, would be more useful.
Both Dr. Tanner and I have complaints about poor and abusive regulatory practices. There is no
denying, however, that some regulations are necessary to protect citizens from poor medical
practices. Government regulation that requires monitoring of skin dose for interventional
procedures is justified.
AGAINST THE PROPOSITION: Raymond L. Tanner, Ph.D.
Opening Statement
Radiation control is achieved by: (a) education and training of users; (b) equipment and facility
design; (c) quality control programs; (d) administrative rules; and (e) government regulation (an
administrative rule). In other words: A qualified expert, using properly designed equipment and
facilities, with continuous quality control, operating within recommended guidelines is an
unbeatable combination for safe use of modern technology. The possibilities of (a), (b), and (c)
should be exhausted before employing (d) and (e) and these must be kept in order.
Development of regulations requiring monitoring of radiation dose during interventional

fluoroscopy may be what is innocently proposed initially. However, this will almost certainly be
quickly followed by extension of the regulations to include specification of maximum dose per
procedure. (Indeed, it might be beneficial to many more patients if a minimum dose(rate) were
required—thus assuring some degree of image quality!) Further extension of dose control to all
diagnostic procedures might well follow as the government agencies evaluate the fluoroscopy
program and proclaim it successful, without thorough independent analysis. State mandated
323
mammographic dose and NRC mandated annual public exposure levels are two examples of
recent unworkable programs. The proposal for interventional fluoroscopy dose regulations
constitutes further undesirable intrusion into the practice of medicine by insufficiently trained
persons blindly following cookbook-like rules. The outstanding recent example of this is
Managed Care which has not only failed to improve the nation's health care but has actually
degraded it.
Technical regulations require that knowledgeable persons be involved both in developing and
implementing them. There are not enough professionals willing to work in the poorly
compensated bureaucratic establishments to provide this. Moreover, the control of radiation dose
in interventional fluoroscopy is possible within existing rules and regulations by means of the
recently lowered maximum allowed exposure rates of such units. This can be followed by
increased education requirements for physicians authorized to use such equipment and by
enhanced quality control programs.
The first quarter of the last century saw the development of fundamental diagnostic radiologic
procedures and the beginnings of radiation risk assessment [mostly very small] from such
procedures. During this period also the international and national radiation advisory groups were
formed. Voluntary programs for radiation control based on the recommendations of these
committees worked well for the second and third quarters of the 20th century. The latter half of
the century saw an influx of government regulations which have obscured the effectiveness of
voluntary programs. Such regulatory mandates are always deemed effective; otherwise the
budgets and positions of the regulators would be in question.
If regulations requiring monitoring of radiation dose in interventional fluoroscopy are developed

then they should include specific performance standards on new fluoroscopy equipment and
periodic testing of such equipment by qualified physicists. Further, such regulations should not
be promulgated by states but by the federal government. Otherwise, uniformity will not exist,
manufacturers will be confused, quality control programs will suffer and costs will rise. The
greater need is one of educating the professional users and the public to better understand how to
safely enjoy the great benefit of ionizing radiation and accept the minor, if any, consequent risk
when it is used knowledgeably and responsibly.
We must not regulate every aspect of our life style lest we become an Orwellian "Big Brother"
society.
Rebuttal
I expected my esteemed colleague to have made the following points: (1) the public interest
demands control of dangerous substances, even when used by professionals, (2) exposure
monitoring will not be done voluntarily, (3) no standard will be adopted unless government
mandated, (4) the risk has been overlooked, (5) voluntary programs cannot enforce compliance.
He alluded to points two and four but chose to emphasize an (undocumented) rise in reported
injuries from interventional fluoroscopy which may well be due to a heightened awareness of the
potential risk involved and/or to increased use of the procedure. Establishment of a rigorous basis
for new regulations (which impinge on the practice of medicine) is required prior to initiation
thereof.
The approximately 100 injuries cited by Dr. Wagner occurred over several years yielding an
incidence of 1 per 100 000 (assuming about 10 million procedures were involved). This is the
324
same magnitude of risks we accept in our daily lives from activities such as driving, work-site
tasks, and leisure pursuits (e.g., boating, skiing, and contact sports). This approach leads to the
much more important topic of resource allocation based on a scientific ranking of risks. Where
and how we use our private, corporate and governmental funds to reduce risks should not be
based on media hype and/or public misunderstanding (fear).
I chose to emphasize the carrots of user education, quality assurance, and equipment design
rather than the stick of government control in speaking to the five points above.
325
8.10. Reference values are de facto regulatory limits for patient

exposures
Mary E. Moore and Priscilla F. Butler

OVERVIEW
Reference values (sometimes called investigational levels) are patient exposure levels that trigger
an institutional response because they are set at the upper limit of expected exposures for patients
undergoing specific procedures. When a reference value is exceeded, institutions are expected to
initiate corrective action to prevent a reoccurrence (see report of the AAPM Radiation Protection
Committee). Hence, some physicists consider reference values to be de facto regulatory limits for
patient exposures. Other physicists disagree with this interpretation. This disagreement in
interpretation is explored in this month's Point/Counterpoint.
Arguing for the Proposition is Mary E. Moore, M.S., M.Ed., DABR, DABMP. Ms. Moore has
both a M.Ed. in Science from Temple University, and a MS in Radiological
Health/Environmental Sciences from Rutgers University. She has been a hospital health
physicist, a medical physicist in radiation oncology, a diagnostic imaging physicist and a system-
wide radiation safety officer. She is the radiation safety officer for the Philadelphia VA Medical
Center. Ms. Moore is a past Treasurer and member of the Board of Directors of the Health
Physics Society, and a member of the American Board of Medical Physics Panel of Examiners
for Medical Health Physics. She continues in her appointments to the NJ Commission on
Radiation Protection (since 1987), and the NJ Radiologic Technology Board of Examiners (since
1991).
Arguing against the Proposition is Priscilla F. Butler, M.S., FAAPM, FACR. Ms. Butler is the
Senior Director of the Breast Imaging Accreditation Programs in the Standards and Accreditation
Department of the American College of Radiology. She is also an adjunct associate professor at
the George Washington University Medical Center. Prior to joining the College she spent over 13
years as a medical physicist and faculty member in the Department of Radiology at the George
Washington University Hospital. Ms. Butler also served as a commissioned officer in the US
Public Health Service (Food and Drug Administration) and participated in the start-up and
conduct of their Breast Exposure: Nationwide Trends (BENT) program. Ms. Butler received her
graduate degree in medical physics at the University of Florida in 1974 and was certified by the
American Board of Radiology in Diagnostic Radiological Physics in 1982.
FOR THE PROPOSITION: Mary E. Moore, M.S., M.Ed., DABR, DABMP
Opening Statement
Are Reference Values "de facto" regulatory limits? The key word in this question is "de facto."
Reference Values are "in effect" regulatory limits with the inherent requirement of mandatory
compliance.
326
Whenever professional organizations publish standards or recommendations for issues not

addressed in state or federal regulations, they become the "accepted rule" or current standard of
practice. For example, prior to the Mammography Quality Standards Act (MQSA),1 the
American College of Radiology's (ACR)2 voluntary mammography accreditation program
instituted a maximum mammography mean glandular dose limit of 0.3 rad. At that time, this
Reference Value was a de facto regulatory limit: If exceeded, corrective action was (and is)
required. The applicant facilities empowered the ACR with the authority to impose this dose, and
other de facto regulatory limits, in order to receive the recognition of accreditation.
Other accreditation bodies such as the Joint Commission on Accreditation of Health

Organizations also have standards that are "de facto" regulatory limits. JCAHO accreditation is
essential for hospitals. If JCAHO were to adopt Reference Values as one of it's required
performance standards, the accredited facilities would consider it as a new regulation and ensure
its implementation.
Current regulations do not establish patient exposure investigation limits. However many
medical physicists use the FDA Center for Devices and Radiological Health's Nationwide
Evaluation of X-Ray Trends (NEXT)3 data as a quality control guide. Measured patient skin
entrance exposures for designated studies are compared with relevant NEXT data to identify
unacceptable patient exposure levels. Establishing trigger levels and requiring corrective action
to prevent reoccurrence is the next step in preventing unnecessary patient exposures. This process
is the essence of the Reference Value program, which also parallels the required response when
state and federal regulations are violated. However, implementing Reference Values does not
require regulatory agency notification.
In summary, voluntary implementation of Reference Values effectively establishes self- imposed

patient exposure limits which are "in reality" regulatory limits.
Rebuttal
The interpretation of reference values as de facto regulatory limits is a perception based on the
effective result of adopting action trigger levels. This interpretation issue is different from the
question of whether reference values should be formalized and incorporated into state and federal
regulations.
The de facto regulatory aspect of voluntarily establishing investigational limits occurs when an
action level is exceeded. The facility is expected to investigate, to identify the cause, and to
implement corrective action that will prevent a reoccurrence. This is the same process followed
when a state or federal regulation is violated. Consequently some medical physicists perceive
reference values as de facto regulatory limits for patient exposures.
Different professional organizations have proposed investigational levels based on expected

patient exposures for specific procedures. Each facility is free to adopt or establish the exposure
levels that they feel are appropriate as reference values. Prior to implementation, each facility
should ensure their exposure limits are adequate for good image quality. While the
appropriateness of a numerical value selected as an investigational limit is a very important issue,
it is different from the question of whether reference values are perceived as de facto regulatory
limits.
327
A significant difference between de facto and state or federal regulations is that the numerical
values of the de facto investigation limits are not cast in bureaucratic stone. Individual facilities
should review, and if necessary, change their investigational limits periodically to ensure they
continue to be appropriate for newly acquired technology.
Voluntary implementation of reference values has the same operational effect as mandatory
federal and state regulations. A limit is set. If it is exceeded, corrective action is required to
prevent a reoccurrence. In effect, the process is the same for both voluntary reference values and
mandatory regulations. Hence, the interpretation of voluntary self-imposed mandates as de facto
regulations is in reality accurate.
AGAINST THE PROPOSITION: Priscilla F. Butler, M.S., FAAPM, FACR
Opening Statement
Reference doses for diagnostic radiology have been discussed internationally (for over 10 years)
by the International Commission on Radiological Protection (ICRP). 4,5,6 Although not
specifically addressed as "reference," in 1978 the US Food and Drug Administration (FDA)
published Radiation Protection Guidance to Federal Agencies for Diagnostic X-rays, including
recommended maximum entrance skin exposures for 10 examinations.7 In more recent years, the
AAPM's Task Group on Reference Values for Diagnostic X-Ray Examinations has drafted a
report for publication; the Conference of Radiation Control Program Directors (CRCPD)
Committee on Quality Assurance in Diagnostic X-ray has drafted a revision to its Patient
Exposure and Dose Guide; and the American College of Radiology (ACR) approved an ACR
Standard for Diagnostic Reference Levels in Medical X-ray Imaging at its annual meeting this
year. Although the terms used for these reference doses differ (e.g., the ICRP and the ACR call
them "reference levels," the AAPM calls them "reference values," and the CRCPD refers to them
as "guides"), there is one unifying theme: these numbers should not be used for regulatory
purposes. To understand why, we must first look at how the numbers will be implemented.
Reference values are established by various organizations for selected examinations and
projections, defined patient sizes and, for automatic exposure controlled (AEC) x-ray equipment,
with specific phantoms to simulate the patient. If the examinations are performed properly,
clinical image quality will be generally acceptable at exposures and doses below the values
defined as reference. Consequently, medical physicists should investigate and identify the causes
of doses that exceed reference values under prescribed conditions. The medical physicist must
consult with the physician and assess his or her image quality needs as actions are taken to
reduce the radiation dose. If the physician determines that there is a clinical need for higher
radiation exposure, then the level in use should remain unchanged.
Many imaging facilities, particularly those not under radiology control, do not require a routine
survey performed by a medical physicist, or may only request a survey that does not include
exposure or dose measurements. Medical physicists should inform their clients of the usefulness
of dose measurements and reference values for their clinical practice and their patients. They
must also be sure to obtain measurements under the same conditions for which the reference
values are defined.
Adopting reference values as regulations by state or federal agencies, or using them as de facto
regulations, could significantly interfere with their successful implementation. The major
328
concern is the potential degradation of image quality that may result if dose reduction measures
are put into place by individuals not trained or experienced in the evaluation of image quality.
Reducing dose without a concurrent assessment of image quality by a physician would be
detrimental to patient care.
Turning reference values into regulations could also hamper the widespread adoption of new x-
ray techniques that may require additional dose. Mammography is a case in point. Since 1988,
the average mean glandular dose for a 4.2 cm breast (as measured by state and FDA inspectors)
has slowly increased from 1.33 (Ref. 8) to 1.76 mGy in 2002. This increase has resulted in
significant image quality improvement primarily due to the use of grids, less noisy film and
higher film optical densities to improve contrast. Under the Mammography Quality Standards
Act (MQSA), the FDA wisely established a regulatory limit of 3.0 mGy.9 There is no need for
mammography to be performed at doses above this level, and facilities will be issued a
noncompliance if it occurs. However, this limit is not a reference value. Should a medical
physicist investigate the reason for an exposure that is just below 3.0 mGy? Absolutely. Could
there be consensus on a mammography reference value that is lower than the regulatory limit?
Most likely. But the value would have to be routinely re- evaluated and revised to allow for
image quality improvement trends. Regulations do not allow for that.
Rebuttal
On October 1, 2002, the ACR Council approved the 2003 ACR Standard for Diagnostic
Reference Levels in Medical X-Ray Imaging. When publishing standards, recommendations or
guidelines, professional organizations carefully explain that their publications are not intended to
serve as the "accepted rule." For example, the ACR includes the following cautionary statement
on each ACR Standard: "The standards of the American College of Radiology (ACR) are not
rules, but are guidelines that attempt to define principals of practice that should generally
produce high- quality radiological care. The physician and medical physicist may modify an
existing standard as determined by the individual patient and available resources." These
standards should not be treated as de facto regulations.
Unfortunately some facilities will erroneously treat reference values as de facto regulatory limits.
The real danger exists that many of these facilities will take the easiest measure possible to cut
patient dose below the reference value without assessing their action's impact on image quality
and patient care. These facilities may not invest the necessary time and resources to research their
entire imaging system in order to determine the true cause of the high doses or to assess the
higher doses' impact on patient care. It is essential that these facilities consult with experienced
medical physicists and diagnostic radiologists to make that determination.
Reference values should be treated as "trigger levels." But, as described by the ICRP and other
organizations, exceeding reference values should trigger an investigation, rather than action to
correct the high dose. Corrective action should only be taken if the radiologist determines that
there will be no reduction in image quality or that the reduction in image quality will not interfere
with appropriate patient care.
Finally, reference values and regulatory dose limits (or accreditation pass/fail limits) serve
different purposes. Patient dose regulations and accreditation pass/fail limits are intended to be
maximum values beyond which there is no demonstrated clinical benefit. Reference values are
investigational levels set to encourage quality improvement. Both can coexist.
329
In summary, reference values are an important tool for improving the quality of radiologic
imaging but should not be treated as de facto regulations.
REFERENCES
1. Mammography Quality Standards Act, Public Law 102-539, October 27, 1992 and Title
21 Code of Federal Regulations, Part 900.12(c).
2. Mammography Quality Control Manual 1999, The American College of Radiology,
Reston, VA 20191-4397.
3. R. Kaczmarek, B. Conway, R. Slayton, and O. Suleiman, "Results of a nationwide survey
of chest radiography: Comparison with results of a previous study," Radiology 215, 891 (2000).
4. International Commission on Radiological Protection (ICRP), Recommendations of the
International Commission on Radiological Protection, Publication 60: Annals of the ICRP 21;
No. 1–3 (Pergamon, Oxford, 1990).
5. International Commission on Radiological Protection (ICRP), Radiological Protection
and Safety in Medicine, Publication 73: Annals of the ICRP 26, No. 2 (Pergamon, Oxford, 1996).
6. International Commission on Radiological Protection (ICRP), Avoidance of Radiation
Injuries from Interventional Procedures, Publication 85: Annals of the ICRP 30, No. 2
(Permagon, Oxford, 2000).
7. Department of Health, Education and Welfare, Food and Drug Administration,
"Radiation protection guidance to federal agencies for diagnostic x-rays," Federal Register
43(22), 4377–4380 (1978).
8. O. H. Suleiman et al., "Mammography in the 1990's," Radiology 210, 345–351 (1999).
9. Department of Health and Human Services, Food and Drug Administration,
"Mammography quality standards; final rule," Federal Register. 68(208), 55852–55994 (1997).
330
8.11. Federally mandated imaging standards (e.g., MQSA) serve the

best interests of patients
John McCrohan and Richard L. Morin
Reproduced from Medical Physics, Vol. 25, No. 6, pp. 811-813, June, 1998
OVERVIEW
Since October 1, 1994, all mammography facilities in the U.S. (except those of the Department
of Veterans Affairs) have been required to be certified by the U.S. Food and Drug
Administration (FDA). This requirement is a consequence of legislation entitled the
Mammography Quality Standards Act of 1992 (MQSA) enacted by the U.S. Congress. The goal
of the legislation is to assure that mammography is safe and reliable, and to allow the detection
of breast cancer in its earliest, most treatable stages. The key features of MQSA are: To operate
lawfully, a mammography facility must be certified by the FDA as providing quality
mammography services. For a facility to be certified, it must be accredited by a federally
approved private nonprofit or state accreditation body. To be accredited, the facility must apply
to an FDA-approved accreditation body; undergo periodic review of its clinical images; have an
annual survey by a medical physicist; and meet federally developed quality standards for
personnel qualifications, equipment quality assurance programs, and record keeping and
reporting. The facility must also undergo an annual inspection conducted by federally trained and
certified federal or state personnel.
The MQSA legislation addresses an issue (breast cancer detection and diagnosis) that is fraught
with emotion, and reflects in part an intense lobbying effort by groups concerned with women‘s
health. Whether or not it represents the most effective path to improvement of breast cancer
detection and diagnosis was controversial at the time of the passage of the legislation, and
remains controversial today. This edition of Point/Counterpoint addresses the controversy.
Arguing for the proposition is John McCrohan. Capt. Mc- Crohan, an officer in the U.S. Public
Health Service since 1974, holds a Masters degree in Radiological Sciences from the University
of Washington. He is the Acting Director of the Division of Mammography Quality and
Radiation Programs in FDA and has had a major role in directing the implementation of the
Mammography Quality Standards Act. Involved in mammography since 1976, Capt. McCrohan
was instrumental in the Nationwide Evaluation of X-ray Trends assessments of mammography in
1985, 1988, and 1992. He has served on numerous mammography committees of the ACR,
NCRP, and the Conference of Radiation Control Program Directors.
Arguing against the proposition is Richard Morin. Richard L. Morin is a Consultant in the
Radiology Department, Mayo Clinic Jacksonville, Professor of Radiologic Physics, and member
of the graduate faculty, Mayo Graduate School. Dr. Morin earned his Ph.D. degree from the
University of Oklahoma in Radiological Sciences in 1980. His current research interests include
computer applications in the radiological sciences with particular emphasis in electronic
imaging, medical imaging detectors, transmission, display, and analysis technologies. He has
been active in the ACR Mammography Accreditation Program and has served on the Standards
and Accreditation and Government Relations Committees of the ACR Commission on Physics
and Radiation Safety.
331
For the Proposition: Capt.John McCrohan
Opening Statement
The federally mandated MQSA imaging standards are clearly in the best interests of patients.
Breast cancer is a life threatening disease and a major source of morbidity and mortality.
Mammography is the only proven means of early detection and the best hope for those at risk
from breast cancer. Therefore the focus on mammography is justified.
Mammography is a difficult imaging technique and one in which high quality is extremely
important in breast cancer detection. A significant investment is necessary to achieve the high
quality required. In 1985, the Nationwide Evaluation of X-ray Trends showed that
mammography image quality was quite variable, 1 demonstrating that the investment necessary to
achieve high quality was not being made consistently. Subsequent standards-setting efforts by
the American College of Radiology (ACR) and state regulatory agencies were only partially
successful at addressing this problem and resulted in a ‗‗patch-work‘‘ of inconsistent
requirements. Federal standards were necessary to assure that all facilities would make the
required investment in quality.
The federal imaging standards mandated by MQSA provide a nationally uniform baseline for
performance, improving image quality, and thus the effectiveness of mammography, particularly
among facilities where image quality was poorest. MQSA standards are risk-based,2 addressing
those aspects of mammography essential for quality. The standards are also performance-based,
allowing flexibility. The standards setting process is open, providing opportunities for the public
and the mammography community, directly and through the mandated advisory committee, to
help define the standards assuring balance between high quality and access, between the benefits
to patients and the burdens on facilities. In fact, the MQSA standards embody, to an
extraordinary extent, the consensus imaging standards developed for the ACR‘s voluntary
Mammography Accreditation Program. As a result the MQSA standards are reasonable,
achievable, and enforceable.
In addition to mandating federal imaging standards, MQSA also requires annual facility
inspection and enforcement of the standards. Both significant effort and the will to invest that
effort is necessary to achieve the quality required by the standards. While initial inspection
results were quite positive [30% of facilities had no findings and 2% had serious findings] and
have continually improved [most recently 55% of facilities had no findings and <1% had serious
findings], they also show that meeting the MQSA standards continues to be difficult for some.
Without inspections many facilities would not make the investment necessary to assure quality.
MQSA had improved quality nationwide without reducing access3 and its federally mandated
imaging standards assure women of high quality mammography. Clearly MQSA is in the best
interests of patients.
Rebuttal
Dr. Morin asks if standards and enforcement should come from the same source. In fact they
must. The federal government cannot enforce nongovernmental standards. The MQSA
‗‗interim‘‘ regulations incorporated standards developed within the radiology community into
comprehensive federally mandated standards. The MQSA ‗‗final‘‘ regulations were developed in
332
an open process involving the National Mammography Quality Assurance Advisory Committee
and 1900 members of the radiology community whose 8000 comments yielded a much improved
federal standard. With MQSA providing a baseline, the radiology community can, and perhaps
should, ‗‗raise the bar,‘‘ developing more stringent standards. However, MQSA standards are
federally enforced, providing a high level of assurance that all facilities meet this baseline.
Dr. Morin states that patients need assurance that their examinations are competently performed
and interpreted. However, neither professional standards nor federally mandated standards assure
competence. What MQSA assures is that all facilities undergo periodic peer review (including
clinical image evaluation) and annual inspection. Through these mechanisms, federally mandated
standards are enforced on all mammography facilities, assuring that equipment performs
appropriately, quality control programs are adequate, personnel meet substantive initial and
continuing qualifications, and examination results are appropriately communicated to patients.
The inspection results [1995, 32% ‗‗no-findings‘‘ inspections; 1997, 56% ‗‗no-findings‘‘
inspections] show both the need for, and the benefits of, inspections.
It is quite true that the best interests of patients are tied to good facilities, personnel, and
practices. However, these can only be encouraged, not assured, by professional standards. Poor
facilities, personnel and practices can only be discouraged, not eliminated. Good facilities can be
assured by appropriately developed federally mandated standards. Patients should not have to
search for a quality mammography facility. They should expect that all mammography facilities
provide quality services.
MQSA is arguably the best example to date of the synergy that is possible between the radiology
community and the government. The results are clearly in the best interests of patients.
REFERENCES
1. B. J. Conway et al., ‗‗Mammography in the Eighties,‘‘ Radiology 177, 335–339 (1990).

2. F. Houn et al., ‗‗The Mammography Quality Standards Act of 1992,‘‘ Radiologic Clinics
of North America 33~6!, 1059–1065 (1995).
3. ‗‗Mammography Services, Impact of Federal Legislation on Quality, Access, and Health
Outcomes,‘‘ GAO Report [GAO/HEHS-98-11, October 1997].
Against the Proposition: Richard L. Morin
Opening Statement
How could anyone oppose federal standards to assure high quality patient care? How could
anyone oppose federal standards to assure all mammography imaging systems are evaluated in
the same way? Interestingly, these are very different questions. The issue is the degree of
involvement of the federal government in the type, nature, and evaluation of imaging standards.
Opposition to standards is not the issue, but rather, who should set the standards and should both
standards as well as enforcement come from the same source.
Actually, in this arena, the standards must be formulated by the regulated community since only
they possess the expertise and training to separate good practice from bad practice. The point
here is that unlike other areas of regulation, specifications of technical factors alone are
necessary but not sufficient. Standards set responsibly by professionals in the field will
333
inevitably be higher than those set by government bodies concerned with regulatory consistency
and inclusion of the least common denominator. Standards set by professional bodies address all
areas of practice and provide the baseline for Continuous Quality Improvement, Critical
Pathways, and other techniques of practice assessment. It is the nature of these standards that
their development involves input from a large number of professionals (e.g., Radiologists and
Physicists) involved in an area (e.g., mammography). The incentive in this setting is competent
patient care, not compliance. Professional standards continue to emerge for Radiology, Radiation
Oncology, and Medical Physics. These standards are becoming more robust and inclusive,
‗‗raising the bar‘‘ in each area of practice.
Let us return to the central point—the best interest of the patient. A patient needs to be assured
that a facility is able to competently perform the required examination and provide a competent
interpretation. The best interest of the patient is tied closely to good facilities, systems, practice,
and physicians. The patient is best served by the best medicine. Hence, for mammography, that
means the best from all aspects of practice: the RIS, the mammography and processing systems,
patient positioning, radiographic technique, image evaluation, interpretation, and communication
to the patient. Standards must therefore involve all aspects of practice.
The federal government has a legitimate role in mandating adherence (and possibly linking
reimbursement) to professional standards. The role of establishing, reviewing, and revising
professional standards belongs with the professionals.
Rebuttal
I fear perhaps a central point was missed. The issue under discussion is ‗‗Federally mandated
imaging standards.‘‘ While the Mammography Quality Standards Act was posed parenthetically
as an example, the issue is certainly much broader, bringing into question the nature and
relationships of agencies involved in standards, regulations, and enforcement. The point here is
not that high quality standards and accreditation programs can save lives; the point is how and by
whom are the standards to be formulated.
However, let us examine the case of mammography. The standards were formulated by the
American College of Radiology with the consultation of a large body of Radiologists, Medical
Physicists, and Technologists. The subsequent accreditation program grew from an early
program administered by the Illinois Radiological Society (a chapter of the ACR). The standards
and programs were developed without regulatory compliance in mind and centered on the
professional competence of personnel and the technical performance of the equipment. These
efforts were focused on the best interest of the patient, in this case, the early detection of breast
cancer. Subsequent legislation and rules certainly provided more prescription in both testing and
compliance. However, it is not self-evident that prescriptive checklists for staff and equipment
improved the prior use of professional judgment. One of the fundamental reasons for the success
of this program lies in the fact that the infrastructure of these standards had been built by a body
independent of the regulatory body. In fact, a strong case can be presented that accreditation
agencies ought to be separate and independent of regulatory agencies. This not only avoids
potential conflict of interest but assures independent professional and patient-centric
though during standard development.
The federal government has a legitimate role in mandating adherence (and possibly linking
reimbursement) to professional standards. The role of establishing, reviewing, and revising
professional standards belong with the professionals.
334
8.12. The AAPM should develop protocols generically, and avoid

documents that are too strict and prescriptive, to facilitate their
adoption by regulatory agencies like the Nuclear Regulatory
Commission and States
Cynthia G. Jones and Bruce Thomadsen
Reproduced from Medical Physics, Vol. 27, No. 3, pp. 431-433, March, 2000
OVERVIEW
Clinical physicists have the most knowledge about specific calibration and quality control
protocols needed to assure the optimized use of ionizing radiation for diagnostic and therapeutic
purposes. Many individuals, including some in regulatory agencies, believe that physicists
should develop these protocols without concern for extraneous influences such as rules and
guidelines from regulatory agencies. However, development of a protocol without concern for
regulations creates an enigma for the practicing medical physicist, because following the
protocol could result in noncompliance with regulations. This enigma could be resolved if
regulatory agencies were willing and able to respond quickly to new protocols, but bureaucracy
of the agencies may interfere even if the agencies wished to respond rapidly. Hence, most
protocols in medical physics are built to encompass regulations rather than to optimize the
procedures they are designed to address. This issue of Point/Counterpoint addresses the intrinsic
conflict between existing regulations and protocol development.
Arguing for the Proposition is Cynthia Jones. Cynthia Jones is the Senior Assistant for Materials,
Office of Commissioner Dicus, Nuclear Regulatory Commission (NRC). Before joining the
Commissioner‘s staff in July 1999, she was the Senior Level Advisor for Health Physics at the
NRC. In that position, she helped develop an NRC Management Directive on the Development
and Use of Consensus Standards. Ms. Jones will receive her Ph.D. in nuclear engineering from
the University of Maryland in September 2001. Before working at the NRC, she was a physicist
at NIST, and a medical and reactor health physicist at UCLA.
Arguing against the Proposition is Bruce Thomadsen, Ph.D. Dr. Thomadsen, a member of the
Medical Physics Department of the University of Wisconsin, specializes in radiotherapy physics
and radiation safety. He has participated in the development of task group recommendations
(compromising as necessary), and with the development of new radiation regulations for the
State of Wisconsin. One of the main thrusts of his work has been in quality achievement and
error prevention in patient treatments.
FOR THE PROPOSITION: Cynthia Jones
Opening Statement
Since the enactment of the National Technology and Transfer Act of 1995 [Public Law ~P.L.!
104-113] on March 7, 1996, all Federal agencies are required to use standards developed by a
consensus body. Although one could argue that the AAPM is not listed as one of the ‗‗official‘‘
standards consensus bodies that are identified on the National Institutes of Standards and
335
Technology website [like the American National Standards Institute (ANSI)], the protocols it
develops use a consensus process within its professional organization. An interesting aspect to
this law is that if a Federal agency uses its own standards (i.e., a regulation developed by an
agency for its own use) in a regulation, instead of using an existing consensus standard, it must
justify the reason for not adopting the standard and provide a yearly report from the head of that
agency to the President‘s Office of Management and Budget (OMB).
In order to effectively communicate how this law was to be implemented at Federal agencies,
OMB issued its revised Circular A-119, ‗‗Federal Participation in the Development and Use of
Voluntary Consensus Standards and Conformity Assessment,‘‘ on February 19, 1998. In that
circular, OMB also requires that when promulgating a proposed rule, a Federal agency must
include a statement that identifies when a voluntary consensus standard is being proposed for
use, or when a government unique standard is proposed instead of a voluntary consensus
standard. In the latter case, the agency must provide a preliminary explanation of why use of a
voluntary consensus standard is inconsistent with applicable law, or is otherwise impractical
(Note: OMB defines ‗‗impractical‘‘ as including circumstance in which use of the consensus
standard would fail to serve the agency‘s program needs; would be infeasible, inadequate,
inefficient, or inconsistent with the agency‘s mission; would impose more burdens, or would be
less useful than the use of another standard). In addition to this statement, Federal agencies must
now also invite public comment as to whether or not the public knows of an existing standard
which should be used in lieu of the proposed regulation.
In developing its proposed regulations since the enactment of P.L.104-113, NRC staff now
consistently reviews the technical literature for determination of any consensus standards or
technical professional societies guidance documents, such as AAPM protocols, that may be used
in lieu of the proposed rule or referenced in a proposed regulation or guidance document.
Considering the subject at hand, if AAPM were to have developed protocols which could be
viewed as suggested recommendations, for example, for calibration and quality control use that
were not in conflict with an existing national consensus standard (such as an ANSI), there would
be incentive indeed for the NRC to use that instead of development of a new rule, or at a
minimum, add that protocol to the list of references which would provide licensees with
guidance in this area. In an era of ‗‗rightsizing‘‘ government, it makes good sense to not reinvent
the wheel, but rather to use already existing professional societies‘ protocols for use in new
regulations.
Rebuttal
The intent of P.L.104-113 and OMB Circular A-119 is to use already existing standards (or
protocols in AAPM‘s case) in lieu of an agency drafting new regulations. If NRC incorporates a
standard or protocol into a regulation in its entirety, it would not be able change any ‗‗shoulds‘‘
to ‗‗shalls,‘‘ because that would require a change to the standard itself. I am not proposing, as the
opposition states, that the AAPM, or any other standards developing organization, write
minimalist standards. Indeed, I am recommending that these bodies more carefully consider what
is essential, versus what is purely optional for licensees to do in the course of a particular
practice. Perhaps wording such as ‗‗Suggested Best Practices‘‘ in protocols would clearly
delineate those ideas that could improve a practice, but yet are nonessential. Without suggested
good practices from societies like AAPM, regulatory agencies would be at a great disadvantage
in having to expend resources to develop their own guidance documents, which is exactly not
what was the intent of P.L. 104-113. Carefully consider what is required of licensees in any
AAPM protocols. Through the use of OMB Circular A-119, you‘ll get what you asked for.
336
AGAINST THE PROPOSITION: Bruce Thomadsen, Ph.D.
Opening Statement
In many AAPM documents we walk a fine line that never quite gets resolved in the organization:
Whether recommendations should describe minimum acceptable standards of practice or
something better. As the technology and our understanding improve, we, as a community, like to
believe that the level of care we can provide our patients also improves. Although the previous
minimum standard usually remains safe and as efficacious as before, we like to suggest that our
members hold their practice to the improved level. We also like our documents to be
comprehensive, covering all possible aspects of a problem.
AAPM protocols for calibration and quality assurance serve two basic functions: (1) to provide
guidance for members performing specific functions, and (2) to improve the state of practice. For
both of these functions, the protocols often contain recommendations considered absolutely
necessary, demarcated by the use of ‗‗shall,‘‘ and other recommendations that would improve
patient care but remain dispensable, indicated by ‗‗shoulds.‘‘ The ‗‗shoulds‘‘ help refine a
practice and often provide layers of safety or assurance in treatments, at the cost of additional
time and resources. Leaving these recommendations optional, the AAPM recognizes that few or
no programs can afford the dedication of personnel or funds to perform all suggestions. Nor do
all suggestions apply to all situations. Compliance with all recommendations in the reports of
Task Groups 40, 53, 56, 59, and 64 alone requires more staffing than reimbursements allow. Yet,
the suggested recommendations frequently help practitioners deal with specific situations, and
avoiding inclusion would leave members without valuable guidance.
Regulatory bodies often cannot use the word ‗‗should‘‘ in rules as the AAPM does. In translating
recommendations into regulations, some agencies simply change all ‗‗shoulds‘‘ to ‗‗shalls,‘‘
adding requirements to perform certain actions uselessly in many situations. Adopting a protocol
en bloc, such as, ‗‗User shall follow AAPM TG#...‘‘ can be interpreted as requiring all precepts.
To avoid this unpleasant consequence of good intentions, the proposition suggests writing
minimalist protocols such that no one could object to performing all parts. Unfortunately, this
approach not only eliminates guidance though suggestions, but also loses opportunities to raise
the quality of patient care. Such an approach also means that no protocol document would collate
all relevant information, leaving members to comb through the literature themselves.
Although possibly uncomfortably increasing the workload for some practicing medical
physicists, the inclusion of some or many protocol recommendations into regulations often
provides the justification for increasing the staffing levels in some programs.
To serve best the AAPM membership and patients under our care, the Association should
continue to draft protocols addressing the state of the art, including necessities and suggested
‗‗niceties.‘‘ During comment periods in regulation formulation, the Association must lobby for
inclusion of the necessary concepts, without adoption of the optional suggestions. To follow the
proposition would set all standards to the lowest common levels.
Rebuttal
337
In an era of ‗‗right-sizing,‘‘ it does make sense not to reinvent anything (or at any other time for
that matter). It would serve the society as a whole well for regulatory bodies to make use of
AAPM protocols. However, simply copying over the protocols into regulations is a case of good
advice making bad laws. One suggestion to avoid mass adoption of AAPM recommendations
into regulations might be to keep official protocols very basic, and move all optional
recommendations into less official formats. In many cases, though, such a procedure would gut
the true substance of the protocol, and certainly make links between the two documents more
difficult. Nor would that prevent agencies from incorporating the less official documents.
Examples of misdirected good intentions abound (such as Minnesota‘s establishing an exposure
limit for the general public of 0.054 mR/y), and some state legislator may think that
incorporation of all the bells and whistles would benefit their constituency.
Thus, with no guarantee of control over the fate of our documents, we could prepare for the
worse, creating only minimalist recommendations that everyone (and their mother or father)
could satisfy. This approach would put the AAPM out of the science business, leaving us as a
self-serving professional (or unprofessional) organization. Or we could continue to write
protocols we consider in the best interest of the patient and society, reflecting the state-of-the-art
and improving the state-of-the-practice, and diligently comment during regulation generation.
All AAPM documents go through extensive review in the Association, from the task group,
through committee and council, and often by the Board and journal reviewers. This process
usually leads to high quality, thoughtful recommendations. We should trust ourselves enough to
continue this course.
338
8.13. NRC restrictions on the packaging of radioactive material

should be expressed more explicitly than simply in terms of
“apparent activity”
Michael S. Gossman and Beth Felinski-Semler
OVERVIEW
As promulgated by the Nuclear Regulatory Commission, packaging regulations for radioactive

material are confusing (e.g., "activity" vs "contained" activity vs "total" activity). As a
consequence, medical physicists are forced to second-guess the intent of the regulations. This
dilemma is the subject of this month's Point/Counterpoint. These authors wish to remind the
readers of Point/Counterpoint that the views presented here do not reflect the views of any
regulatory body mentioned.
Arguing for the proposition is Michael S. Gossman, M.S. Beginning academically at Indiana
University, he furthered his education by attaining a Master's Degree at the University of
Louisville. Pioneering in atomic physics, he eventually published a book on scanning tunneling
microscopy in 1997. Mr. Gossman attended Vanderbilt University, where he studied medical
physics and worked as a health physicist. His current focus is research and planning for special
procedures in the area of high dose-rate brachytherapy. Mr. Gossman is certified in Therapeutic
Radiologic Physics by the American Board of Radiology (ABR) and is a clinical medical
physicist at Erlanger Medical Center in Chattanooga, TN.
Arguing against the Proposition is Beth Felinski-Semler, M.Sc. She received her B.A. in Physics
from LaSalle College (University) in 1976, and her M.Sc. from Temple University in 1977. She
began working in medical physics in 1977 at Cooper Hospital/University Medical Center in
Camden, New Jersey, first in nuclear medicine and later in radiation therapy. She has instructed
both radiation oncology residents and therapists in physics and dosimetry. She is certified in
radiation therapy physics by the ABR. Beth is presently the senior clinical radiation therapy
physicist at the Department of Radiation Oncology, South Jersey RMC in Vineland, New Jersey.
FOR THE PROPOSITION: Michael S. Gossman, MS, DABR
Opening Statement
There are troubling inconsistencies in the practice of transporting sealed radioactive sources in
the United States. There are currently no uniform guidelines governing the proper disclosure of
radioactive sources in transit. The purpose of such disclosure is to enable both receivers and re-
sponse teams to properly assess situations involving damage to packages containing radioactive
material. Unfortunately, physicists are left to use their own interpretations as to the method for
labeling such packages and determining which values to disclose. Without specific guidelines,
many questions remain unanswered. Notably, which activity level should be used for labeling
packages—the apparent activity or the contained activity? In many instances there is a great
339
disparity between the two. Currently, a survey including twenty brachytherapy source
manufacturers revealed half are labeling according to the apparent activity rather than the
contained activity.1 My discussions with some of our collegial society members on how they
transport radioactive materials validate the breadth of the problem.
Consider, for example, the inconsistencies in the shipment of prostate seeds. The (contained
activity to apparent activity) conversion factors for all prostate seed manufacturers range from
1.30 to 1.78 for iodine-125 and from 1.80 to 2.20 for palladium-103. Depending on which
activity the physicist chooses for labeling, the value could differ from what another might label
by as much as 78% for iodine-125 and 220% for palladium-103. There is a potential for a source
encapsulation to break open and leave a bare source with a higher exposure rate than that which
would occur if the encapsulation did not break. Currently, response personnel are not equipped
with enough information to know the worst-case scenario when radioactive materials are
damaged. This deficit could have devastating results.
Clearly we need uniformity to ensure all avenues of safety. Regulations governing the shipment
of radioactive material, however, do not specifically address this troubling issue. No guidance
regarding proper labeling criteria is available from the Nuclear Regulatory Commission, the U.S.
Department of Transportation, the U.S. Department of Energy, the International Atomic Energy
Agency or the International Air Transport Association.2,3,4,5 Explicit guidelines need to be
promulgated and followed to govern the process for labeling radioactive material. To do so, the
conversion factors for apparent to contained activity need to be determined and published for all
sealed sources. Using this information, a guideline should be produced by the Nuclear Regulatory
Commission to indicate what the explicit labeling standard will be. Furthermore, it should be
uniformly recognized by the other supervisory agencies and departments.
Rebuttal
Modern treatment planning systems use air-kerma strength and other appropriate factors
inclusive of apparent activity for source specification.6,7 For older models, source specifications
come from the activity and gamma factor stated by the vendor. As my worthy antagonist affirms,
some vendors maintain they do not indicate whether the activity is "contained" or "apparent," and
some do not even provide the conversion factor.
The FDA requires that manufacturers of radiological devices specify how the source
encapsulation (and substrate) influences the output.8 The physical quantity of activity is also to be
listed as "apparent" or "contained."8 These data were not required for sources made long ago.
Still, conversion factors can be determined experimentally. The method has been published, and
results were provided for sources currently used in intravascular brachytherapy treatments.9,10
The method involves assaying the output of the encapsulated source and then assaying the
material again when the encasement has been chemically digested with an acid.
Guidelines from no regulatory body specify to label radioactive material shipments according to
apparent or contained activity. This implicit wording is nothing short of permission to do either.
It is my suggestion to have regulations explicitly state "apparent activity" when the activity rating
in shipments is used. Furthermore, it is my suggestion that the shipper make the conversion
factor for that source model readily available by presenting it in the clear pouch external to the
package.
340
To attain uniformity, the NRC should first require manufacturers to provide the conversion factor
for each source model. This information should be available, since it was originally requested by
the FDA for medical use approval. For sources that are no longer available, documentation may
be obtained from the FDA or in the files of the applicable former manufacturer. If factors are not
available, they can be identified using the method discussed previously. Moreover, the factors
must be made available before these proposed regulation changes are introduced.
I agree that such a change in shipping regulations will affect the wording in other regulations. It
will affect and should affect regulations like those from the IATA internationally. Regardless of
the inconvenience associated with change, we need to endorse strictness and uniformity. Only by
establishing a standardized method for classifying and monitoring radioactive material, can we
appropriately account for what is transported.
AGAINST THE PROPOSITION: Beth Felinski-Semler, MSc, DABR
Opening Statement
In the field of medical physics, definition and consistency of procedures have always been
important. These are the foundation for our field today. We improve and make things better and
clearer, but are always able to trace back to the foundation. We now have an inconsistency in the
shipping and receiving of brachytherapy "sealed sources." The confusion occurs because of the
specification of source strength. Source strength has been defined in one of four ways: the
contained activity, the apparent activity, the equivalent mass of radium, and the exposure rate at a
specified distance.11 Three of these methods depend on the inherent filtration associated with
source construction. It is this filtration that gives rise to the term "apparent activity" for clinical
use, because the apparent activity is less than the contained (actual) activity due to source
filtration. The exposure to persons handling the sources, and the dose to the patient, are not
dependent on the actual (contained) isotope activity, but instead on the "filtered" activity outside
of the source walls. Hence the problem: which activity, apparent or contained, should be used
when describing sealed sources in transit.
The NRC (Part 20 appendix A) establishes activity levels and their shipping requirements for
specific sources. In Part 71 the NRC defines and establishes shipping container requirements.
Then the DOT and IAEA take over and establish labeling requirements based on the exposure
level at the surface of the package and at one (1) meter. The purpose of all these regulations is to
set universally known safety standards. Are any of these regulations affected if the apparent
activity or the contained activity is used, even if in some cases such as palladium-103 the
difference could be as much as 200%? I think not. The difference in activity only comes into
consideration if a source is ruptured in some type of accident. In this situation, one would hope
the outer shipping container that the sources are housed in will still be adequate to contain the
activity at the required level. If contamination is the concern, first and foremost to the emergency
response personnel is the qualitative knowledge of what the exposure rate in the area is, then
what the specific isotope is, what type of radiation is involved, and what the physical form is.
What the labeled activity states is not the primary concern in this situation.
So, should the use of apparent or actual activity be causing so much concern? No! Is there a
problem with not establishing which of the two should be used? Yes! As stated previously,
consistency has always been a mainstay for our field. Therefore the answer is simple: we should
use apparent activity. The source has inherent filtration which is an intrinsic part of the source
341
itself. The filtration is not removed when the source is used clinically. All clinical
documentation, such as the written directive, describes the source in terms of apparent activity.
All computational systems use apparent activity.
In closing I have one last thought. If it were to be decided that contained activity should be the
standard for shipping, then it will be necessary to alter the departmental paperwork to state this
activity. Who will supply the factors to convert all of the available sources? What about long-
term existing sources whose manufacturer no longer exists? I have called several of the
companies who supply seed sources. Several of them do not supply this information.
Rebuttal
I agree with my colleague's desire for uniformity, as evidenced in my opening statement—

definition and consistency are a foundation for our field. But I do not believe that if actual
activity is not used, we are putting ourselves and emergency response personnel at risk. Have
there been hospital incidents where seeds have ruptured? Yes, mostly due to errors in judgment.
With proper handling, these incidents would not create dangerous exposure levels. "Time,
distance, shielding and containment" are the keys to radiation safety. Policies and procedures are
in place in all institutions using radioactive material; we just have to read them!
What about the public arena? There are policies and procedures there also. The federal
government and individual states have emergency response policies, a section of which covers
radiation emergencies. In my home state of New Jersey, the state12 has a radiation response team
under the jurisdiction of the State Police. There are two levels of response: an awareness group
and the HAZMAT technicians. The awareness group is the first responder to a site. They know
how to evaluate and recognize a hazardous situation, and they have responsibility for defensive
measures—the protection of life and property including evacuation of an area and its security
when called for. They also are responsible for notifying the HAZMAT team when warranted.
HAZMAT is responsible for measuring, containing, and removing dangerous materials. In case
of an accident involving radioactivity, the exposure level at the site is the controlling factor.
My knowledgeable opponent states that if actual activity is to be used, the conversion factors
from apparent to actual activity need to be supplied. We are of like minds here: I just posed the
question "Who" should supply the information. This is no trivial matter. The easy answer would
be the NRC. They have written the guidelines for activity levels and shipping requirements, and
they license the manufacturing of sources and their use in institutions. But this is true only for
reactor-made products. What about accelerator-produced isotopes? For these materials the States
are in control. In addition we must not leave out the DOT. This agency has definitions and areas
of control of their own, which may agree with the NRC and States, but are not limited by them.
These are only three of the multiple groups who are responsible for defining activity. All of these
groups will have to be involved. So perhaps my simple question "Who should supply the
information we need?" should be rephrased to, "When will this issue be resolved?"
I agree that a guideline is needed for all to follow, when describing the activity of a sealed source.
We simply cannot continue to use both actual and apparent activity. I doubt that much will
happen soon. In the meantime, the use of apparent activity is the best way to proceed when
dealing with sealed sources. In the event of an accident, we can put our minds at ease by
knowing that the exposure level is the controlling factor in how the event is dealt with, not a
number written on a slip of paper.
342
I wish to thank Daniel Januseske, M.S. for the information and time he shared with me during the
past month.
REFERENCES
1. M.S. Gossman, Survey-Shipping methodology of current brachytherapy source

manufacturers (2004).
2. US Code of Federal Regulations, Title 10, Energy, Nuclear Regulatory Commission
(NRC) and U.S. Department of Energy (DOE), cur. ed.
3. US Code of Federal Regulations, Title 49, Transportation, U.S. Department of
Transportation (DOT), cur. ed.
4. International Atomic Energy Agency (IAEA), Safety Standards, Regulations for the Safe
Transport of Radioactive Materials, Safety Series 6, 7, 37, and 80, cur. ed.
5. International Air Transport Association (IATA), Dangerous Goods Regulations, 44th ed.
(2003).
6. Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43,
"Dosimetry of interstitial brachytherapy sources," Med. Phys. 22, 209–234 (1995).
7. Update of AAPM Task Group No. 43. Report, "A revised AAPM protocol for
brachytherapy dose calculations," Med. Phys. 31, 633–674 (2004).
8. US Food and Drug Administration (US FDA), Department of Health and Human
Services, Center for Devices and Radiological Health, Device Evaluation Information, Guidance
for the Submission of Premarket Notifications for Photon-Emitting Brachytherapy Sources
(2002).
9. R. Colle, "Chemical digestion and radionuclide assay of TiNi-encapsulated 32P
intravascular brachytherapy sources," Appl. Radiat. Isot. 50, 811–833 (1999).
10. R. Colle, "Activity characterization of pure beta-emitting brachytherapy sources," Appl.
Radiat. Isot. 56, 331–336 (2002).
11. F.M. Khan, The Physics of Radiation Therapy (Williams & Wilkins, 1984).
12. New Jersey State Emergency Response Plan,
http://www.nj.gov/dep/infofinder/topics/emergency.htm (accessed 5/2004).
343
8.14. In addition to the current byproduct materials, it is important

that the Nuclear Regulatory Commission take over regulation of
naturally occurring and accelerator-produced radioactive materials
Kevin L. Nelson and J. Frank Wilson
OVERVIEW
Senator Hillary Rodham Clinton recently introduced a bill into the US Senate (Bill No. S.2763)
to " amend the Atomic Energy Act of 1954 to redefine "byproduct material" to include: (1)
any discrete source of Ra-226 produced, extracted, or converted after extraction, for use in a
commercial, medical, or research activity; (2) any material that has been made radioactive by
use of a particle accelerator and is produced, extracted or converted after extraction for use in
such activities; and (3) any discrete source of naturally occurring radioactive material, other
than source material, that the NRC determines would pose a threat similar to that posed by a
discrete source of Ra-226 and is likewise extracted or converted after extraction, for use in such
activities." The bill also instructs the NRC to promulgate final regulations and standards to
achieve this and to cooperate with States in formulating such regulations.
This is a controversial issue, especially as regards the medical use of these radioactive materials,
when one considers that, in 1996, the National Academy of Science's Institute of Medicine
proposed to Congress1 that " Congress eliminate all aspects of the NRC's Medical Use
Program, 10 CFR Part 35, and those regulatory activities conducted under 10 CFR Part 20 that
are applicable to medical uses". The assumption of regulation of all radioactive materials by the
NRC is the subject of this month's Point/Counterpoint debate.
Arguing for the Proposition is Kevin L. Nelson, Ph.D. Dr. Nelson received his Ph.D. from the
University of Minnesota in health physics. He has been employed as a medical health physicist at
Mayo Clinic in Jacksonville, Florida since 1995 and is an Assistant Professor of Radiology in the
Mayo Clinic College of Medicine. He is certified by the American Board of Health Physics. Dr.
Nelson has served on the Board of Directors of the Health Physics Society from 2002–2005. He
also served as an Associate Editor of the Health Physics Journal for eleven years.
Arguing against the Proposition is J. Frank Wilson, M.D. Dr. Wilson is Professor and Chairman
of the Department of Radiation Oncology and Director Emeritus of the Cancer Center of the
Medical College of Wisconsin. He has served as the President and Chairman of the Board of the
American Society for Therapeutic Radiology and Oncology, and as President of the American
Radium Society. In addition, he has served as Chancellor and as Vice President of the American
College of Radiology. He is the PI of the ACR Quality Research in Radiation Oncology (Q-
RRO) project. Dr. Wilson is a member of the National Council for Radiation Protection (NCRP).
FOR THE PROPOSITION: Kevin Nelson, Ph.D.
Opening Statement
344
Should the oversight of accelerator-produced radioactive material and certain naturally occurring
radioactive material (NARM) be unified? I would argue that centralized control is better, even
though, in the past, these materials have been under some form of federal agency oversight or
state control. To better answer this question, it is useful to review the history of regulatory
oversight of radioactive material in the United States. In 1946, the Atomic Energy Commission
was created as part of the Atomic Energy Act.2 No mention was made of accelerator-produced or
naturally occurring radioisotopes. With time, demand for these isotopes increased due, in large
extent, to the interest in medical applications. With the advent of this technology, standards were
required for both reactor and accelerator-produced radioactive material. In the late 1950's–1960's,
the U.S. Department of Commerce issued a series of radiation standards. Accelerator produced
materials, including 18F, were included.3 Since the 1970's individual states have been allowed,
pending approval from the Nuclear Regulatory Commission (NRC), to create regulatory
programs compatible with federal regulatory standards. To date, thirty-three such agreement state
programs exist,4 all with slightly different regulatory wording. Most agreement states, however,
regulate accelerator-produced materials and naturally occurring radioisotopes using the same
guidance given to NRC-regulated materials.
There is concern that additional regulations may be required to adequately safeguard all
radioactive material, including generally licensed devices and that additional security measures
are necessary to ensure these materials do not fall into the wrong hands. The Department of
Homeland Security is tasked with providing centralized control and more efficient and effective
regulatory processes to better safeguard this country.5 It is in this context that Congress has
attempted, since 2002, to pass legislation to address the unification of regulatory control over
NARM materials. Language in the congressional bills introduced in the 108th Congress, S. 1043
and S. 2763, addressed three types of materials: (a) 226Ra (b) other naturally occurring radioactive
material viewed to pose a similar threat to 226Ra, and, (c) accelerator-produced radioactive
materials.6 Of course, if similar legislation passes, details regarding compatibility with existing
regulations will need to be resolved. Allowing for proper disposal of NARM material under
existing legislative and regulatory mandates would need to be addressed. The Organization of
Agreement States and the Health Physics Society have issued a joint position statement in
support of this initiative.7
Some may argue that the existing level of regulation is sufficient. However, the United States is
one of the few nations in the world that does not regulate NARM from a centralized perspective.
Under centralized regulation current agreement states would be minimally impacted since they
already regulate accelerator-produced radioactive material and NARM in a similar fashion as
NRC-regulated material.
I agree with remarks made by NRC Commissioner Diaz before the American Radiation Safety
Conference and Exposition on June 11, 2001, "One of the fundamental reasons to have regulation
is to decrease the uncertainty in the implementation of a nation's interests, without undue burden
to society."8
Centralizing regulation of these materials is appropriate.
AGAINST THE PROPOSITION: J. Frank Wilson, M.D.
Opening Statement
345
My position against the proposition serves the dual imperatives of maintaining optimal medical
care and unfettered biomedical research in the future. Expanding the NRC regulatory mandate
risks moving in the opposite direction. In fact, all aspects of the NRC's Medical Use Program 10
CFR Part 35, and those regulatory activities conducted under 10 CFR Part 20, as applicable to
medical uses, should be eliminated, as was recommended to Congress in 1996.1 Intervening
world events since that time, and growing awareness and concern about potential nuclear
terrorism, do not mitigate against such a restructuring. These unfortunate developments, on the
contrary, argue emphatically for a more effective, cost efficient, alternative regulatory system for
the low risk situations in which ionizing radiation is used for medical purposes; not for
propagating the old one. Establishing an overarching regulatory framework that considers all
radiation sources used in medicine in their entirety deserves a high priority and broad based
support. This would be the best way to achieve reasonableness in regulatory applications so that
medical progress is not impeded while, at the same time, a maximum level of public safety is
ensured.
Currently, regulatory authority over the medical uses of ionizing radiation is widely distributed
among numerous governmental agencies at federal, state and local levels.9 These entities and
their complex roles and relationships need not be detailed here. While it can be argued that this
"system" somehow seems to "work," it is cumbersome and difficult to cope with. Inconsistent
regulatory requirements may contain security loopholes that go unrecognized. Expanding
unequal treatment of the radiation sources used in medicine, might aggravate the existing
situation. In addition, resources that would be expended to develop and promulgate a new federal
program, and especially later by medical and research facilities trying to comply with it, would
not yield equivalent benefit in terms of patient protection or increased national security. This
time and money would be better directed to enhancing medical care and research, including the
development of new radiopharmaceuticals.
With specific regard to regulation of naturally occurring radioactive materials, Ra-226 and its
derivatives are surely the most prevalent and dangerous NORM that could be used
malevolently.10 Planning long term regulation of a substance that outlived its usefulness, is
clinically obsolete, and has no significant future role in biomedical research is not necessary. 11
Needed instead is a high profile national action plan to rapidly identify and finally dispose of all
remaining radium sources. Built-in rewards and incentives, both positive and negative, would
gain wide cooperation with the plan. A worldwide effort of this sort should be undertaken. In the
USA, the states, with the leadership of the Conference of Radiation Control Program Directors,
are best prepared to organize and carry out this task. They are those closest to and the most
knowledgeable of possible remaining repositories within each locale of these dangerous
materials.
I look forward to responding to the viewpoint expressed by Dr. Nelson favoring the proposition
under consideration.
Rebuttal: Kevin Nelson, Ph.D.
Dr. Wilson and I agree on the central point of this discussion—inconsistent regulatory
requirements can create problems. The Health Physics Society, for example, has expressed a
similar concern as far back as 1992 when a position statement on this issue was first created.12
Dr. Wilson has eloquently discussed the theoretical detriment to medical and research activities if
an overarching regulatory framework were to be created. The NRC and agreement states have
increasingly attempted to work with key stakeholders on legislation that may impact them. Under
346
this working relationship, I am of the belief that with input from state programs that already
regulate these materials, the final result could truly benefit the end user. I would also agree that
most radioactive material used in medicine and research poses a low risk according to most
international and national radiation safety experts. However, it is important to remember that the
intent of terrorism involving radiation is not so much radiological as it is psychological. Not
regulating NARM and certain naturally occurring radioactive materials uniformly above certain
exempt quantities is, in itself inconsistent, as these materials could cause harm if used for
nefarious purposes. In addition, creating a unified regulatory structure would also benefit non-
Agreement State medical licensees who must now navigate a regulatory quagmire and often pay
a fee to the state for licensing NARM and to the NRC for licensing by-product material. I would
also agree with Dr. Wilson that having a national plan to dispose of all unused or unwanted
radium sources would be beneficial. However, given the fact that we continue to struggle with
the disposal of low-level waste in this country, even after the passage of the Low-Level
Radioactive Waste Compact Act of 1980, I am uncertain when such legislation would be passed.
In the meantime, these sources still pose a risk and need to be properly controlled. I believe this
initiative is important in maintaining regulatory consistency and the benefits will far outweigh
the perceived burden.
Rebuttal: J. Frank Wilson, M.D.
Apparently, Dr. Nelson and I completely agree on key points that surface when considering the
Proposition. First, is the paramount obligation to provide for the national security. Other
agreement areas are as follows: a) uniform regulatory control of sources of ionizing radiation is
prudent; b) the existing regulatory framework is highly complex and may contain unreconciled
incompatibilities; c) proper disposal of NARM material should be definitively addressed.
Where, if at all, do we disagree? First, asserting federal control over NARM does not create a
new alternative, non NRC-based, regulatory framework encompassing all radiation sources used
in medicine. Such a framework would, I believe, be a more coherent, cost efficient system than
that proposed, which would avoid overregulation of low risk medical and research situations.
Enlarging the NRC mandate to achieve security objectives without imposing new administrative
burden and costs on healthcare and biomedical research downstream seems unlikely. New
unfunded mandates imposed on these vital activities can no longer be afforded. Whether public
concern is for protection against "nuclear weapons" or "dirty bombs," federal agencies under
mandate must carry out enforcement to a zero risk level. Many recall that until quite recently
reactor produced radioactive materials used in medicine were regulated at the same intensity
level as nuclear plants themselves and the severe consequences of noncompliance.
I had read the joint statement of the Organization of Agreement States and the Health Physics
Society, as an individual never involved in either organization. It struck me that, although new
legislation is endorsed, the sponsors also indicate seven conditional "principles for enactment"
reflecting background concerns similar to those I expressed. Some of these principles are
mentioned, but not addressed in detail, in pending legislation. At this writing, the legislative bill
has survived Joint Conference Committee review. With Congressional endorsement presumably
imminent, further preliminary discussion of these issues will be moot.
REFERENCES
1. Radiation in Medicine A Need for Regulatory Reform, National Academy Press,

Washington, D.C., 1996.
347
2. Senate Special Committee on Atomic Energy, Atomic Energy Act of 1946. Hearings on S.
1717 January 22-April 4, 1946, 1–9. Washington, D.C., (1946),
http://nuclearhistory.tripod.com/secondarypages/aeact.html
3. U. S. Department of Commerce, Maximum permissible body burdens and maximum
permissible concentrations of radionuclides in air and in water for occupational exposure,
National Bureau of Standards Handbook 69 (U. S. Government Printing Office, Washington,
D.C., June 5, 1959).
4. Organization of Agreement States (OAS) website http://www.agreementstates.org/
5. Department of Homeland Security (DHS) website
http://www.dhs.gov/dhspublic/index.jsp
6. Thomas website http://thomas.loc.gov/home/c108query.html
7. Health Physics News; 33, 15–16 (2005) and
http://hps.org/hpspublications/papers.html#position
8. Nils J. Diaz, "Relevance of radiation protection; remarks before the 46th annual meeting
of the Health Physics Society," Cleveland, Ohio, June 11, 2001. Health Physics Society website
http://hps.org/documents/diazspeech.pdf
9. Who Regulates Radioactive Materials and Radioactive Exposure? U.S. Nuclear
Regulatory Commission, http://www.nrc.gov/what-we-do/radiation/reg-matls.html
10. J. O. Lubenau, A Century's Challenges, Historical Overview of Radiation Sources in the
USA, IAEA Bulletin, 41, (1999).
11. Treatment of Accelerator-Produced and other Radioactive Material as Byproduct
Material, S. 864 Nuclear Safety and Security Act of 2005 (Reported in Senate), Sec. 202,
http://www.theorator.com/bills109/s864.html
12. Health Physics Society, Compatibility in radiation-safety regulations, Position statement
adopted January 1992, reaffirmed March 2001, Health Physics Society website
http://hps.org/hpspublications/papers.html#position
348
8.15. The current NRC definitions of therapy misadministration are

vague, do not reflect the norms of clinical practice, and should be
rewritten
Howard Amols and Jeffrey F. Williamson
OVERVIEW
Application of the concept of therapy misadministration, as interpreted by the Nuclear

Regulatory Commission, has confused many medical physicists. This confusion has caused some
physicists to rail against the definition, and to declare that the concept is unworkable in specific
clinical situations. Others believe that recent changes in the interpretation have made the concept
of therapy misadministration manageable in the clinical setting. This difference in perspective is
examined in this month's Point/Counterpoint.
Arguing for the Proposition is Howard Amols, Ph.D. Dr. Amols received his Ph.D. in Nuclear
Physics from Brown University in 1974, followed by post-doctoral training at Los Alamos
National Laboratory. He has held medical physics positions at the University of New Mexico,
Brown, and Columbia Universities. He is currently Chief of the Clinical Physics Service at
Memorial Sloan Kettering Cancer in New York City, where he dutifully reports a few
misadministrations every year, his personal reservations on the regulations notwithstanding. He
is certified by the ABMP, and is a Fellow of the AAPM. He has over 100 serious publications in
addition to his numerous infuriating letters and editorials. Although Dr. Amols recently became
AAPM President Elect, the views expressed here do not represent those of the AAPM or any
other reputable society.
Arguing against the Proposition is Jeffrey F. Williamson, Ph.D. He is currently Director of

Medical Physics and Professor of Radiation Oncology at Virginia Commonwealth
University/Medical College of Virginia Hospitals. While working as a dosimetrist, Dr.
Williamson completed his Ph.D. in Biophysical Sciences at the University of Minnesota. He has
been a member of the faculty of the University of Minnesota and Arizona and of Washington
University. He is a member of the NRC's Advisory Committee on the Medical Uses of Isotope
(ACMUI), Immediate Past Chair of the Radiation Therapy Committee (RTC), and Chair of
RTC's Subcommittee on Photon-Emitting Brachytherapy Dosimetry. Dr. Williamson wishes to
remind "Point-Counterpoint" readers that the views presented here are his alone and do not
represent those of the ACMUI or the U.S. NRC.
FOR THE PROPOSITION: Howard Amols, Ph.D.
Opening Statement
349
According to most regulatory agencies a medical misadministration1 in radiation therapy means

(give or take a few words from state to state) the administration of "a therapeutic radiation dose
such that errors in computation, calibration, time of exposure, source activity or placement,
treatment geometry or equipment malfunction result in a delivered total treatment dose differing
from the prescribed total dose by more than 10%; or a single fraction dose differing from the
dose ordered for that fraction by more than 50%." Further, "when a misadministration is
discovered the licensee or registrant shall immediately investigate the cause and take corrective
action."
I believe this definition is unrealistic because: (1) The threshold values (10 and 50%) are
arbitrary, because they are unrelated to expectations of significant clinical consequences, and
they are not based on any analysis of whether such errors reflect the standard of care (i.e., do
"good" clinicians rarely make such errors while "bad" clinicians do?). (2) The definitions are
vague. What, for example, does "error in source placement" or "treatment geometry" mean?
Does: "prescribed dose" refer only to the target volume, and if so what fraction of the target
volume. Are incorrect normal tissue doses also misadministrations? Is a 5 mm error in collimator
setting a misadministration if it slightly underdoses a sliver of the target and/or overdoses a sliver
of the spinal cord? For an I-125 seed implant to the prostate, does a single seed implanted in the
rectum constitute a misadministration? Space does not permit a complete list of such ambiguities,
but we leave it to the reader (as a homework exercise) to come up with five similar examples. (3)
For most misadministrations, which in my experience result from garden variety carelessness, the
edict to "take corrective action" often results in little more than banal busywork. What kinds of
corrective action are expected, for example, when a misadministration results from someone
punching the wrong number into a calculator or computer program? Is "our staff has been
instructed to be more careful" an adequate response? Does such action really accomplish
anything?
In a recent editorial in the journal Brachytherapy2 it was suggested that a more logical definition
of a misadministration is to identify only events where, because of some error or lapse in
judgment by the medical team, the dose delivered, or volume irradiated is outside the norm of
acceptable clinical practice and/or likely to cause clinical harm. Revising current definitions
along these lines would require: (1) A survey or analysis to determine the distribution of
treatment errors. Are 10% and 50% dose errors within or outside the normal variation of
delivered-vs-prescribed patient doses when practiced in a "reasonable manner" by "careful and
experienced practitioners"? (2) A survey or analysis to determine if a consensus exists regarding
a dose threshold for adverse clinical consequences to the patient.
Further, we suspect (but cannot prove) that: (1) Misadministrations occur far more frequently
than is reported, due in part to the vagueness of the definitions, and to the fact that many
clinicians fail to see their utility. (2) There is little consistency among institutions in this country
on the interpretation of a misadministration, and whether or not certain types of errors meet or do
not meet the definition of a misadministration.
Finally, a law that cannot be enforced, much less understood, generates only contempt for the
law. Until the rules for misadministration are modified to reflect clinical reality they will remain
ineffective.
Rebuttal
350
Dr. Williamson states that the definition of a Medical Event (ME) must be " decidable
unambiguous and relevant." I agree, but maintain that current NRC definitions of
ME fail on all counts. I've yet to see any evidence suggesting that the current dose threshold
values are anything more than numbers picked at random, nor am I convinced that even the new
definitions of ME are unambiguous. The recent revisions of CFR 35 helped, particularly for
brachytherapy, but still do not set unambiguous criteria for such things as error in source
placement, fraction of target volume that must be correctly irradiated, etc. Dr. Williamson limited
his discussion to brachytherapy, but did not discuss NRC rules for teletherapy ME which are
even more confusing. I will return to brachytherapy shortly, but first would like to remind readers
that NRC ME rules still apply to Co-60 teletherapy, and that these rules have filtered down in one
form or another to state regulations for linear accelerator teletherapy, even though this
technology is not directly regulated by the NRC. Much of my opening statement was focused on
the particularly ambiguous definitions of ME for teletherapy. Since Dr. Williamson did not
address teletherapy, I will assume that he agrees with me.
So let me return to brachytherapy. Dr. Williamson correctly argues that a major purpose of
having ME regulations is to identify programs that have seriously flawed QA programs. But how
can one quantitatively determine whether a QA program is seriously flawed? If a center performs
100 brachytherapy procedures per year, and implants an incorrectly-calibrated seed in one patient
during the course of the year that results in a 20% dose error to one patient, is this a flawed QA
program? How about another program, also doing 100 brachytherapy procedures per year that
discovers a systematic error that resulted in all 100 of their patients receiving a 9% dose error?
Which, if either of these program is seriously flawed? Current regulations say the former, with
the latter program getting off free. I disagree!
Finally, Dr. Williamson challenges me to review his list of FY 2002 NMED ME and point out
recently reported MEs that did not warrant investigation. Space is limited so I'll focus on reported
MEs for intravascular brachytherapy (IVB), of which there were 10. I note first that prescribed
doses for these procedures ranged from 8 to 23 Gy, and that in 5 of these procedures the
delivered doses to the prescribed volumes were in fact within the same range, although they
differed by more than 20% from the prescribed dose for the particular patient being treated. This
is admittedly a simplistic argument that ignores different prescription doses for different
isotopes; the point is that errors in delivered doses for these MEs were of the same magnitude as
the uncertainty in what optimal prescription doses should actually be. In other words we're
punishing the occasional klutz but not the perpetually ignorant. Second, I note that in most of
these MEs, device failure played a significant role, and devices are regulated by the FDA, not the
NRC. So if we really want to do this right, there also needs to be better coordination in
Washington (what else is new?).
In conclusion let me state that I am not against having rules, watchdogs, or minimum acceptable
standards. Quite the contrary! I'm simply arguing that the current NRC regulations need to be
improved.
AGAINST THE PROPOSITION: Jeffrey Williamson, Ph.D.
Opening Statement
The misadministration or medical event (ME) reporting requirement is intended to capture

avoidable technical errors on the part of the caregiver that result in erroneous or unintended
351
delivery of radiation to patients or human subjects. Premature termination of an implant by the

patient, seed migration, or other factors beyond the control of the caregiver that give rise to dose
delivery errors are neither misadministrations nor medical events. The U.S. Nuclear Regulatory
Commission (NRC) has recently replaced "recordable event" and "misadministration" with the
revised concept of "medical event." An ME (10 CFR 35.3045(a)),3 is any event, excluding
patient intervention, in which administration of byproduct material or radiation therefrom that
yields a dose (a) that differs from the prescribed dose by at least 50 cSv to an organ; and a total
delivered dose differing from the prescribed dose by 20%, or a dose delivered by a single fraction
differing from the prescribed dose/fraction by 50%; (b) that exceeds 50 cSv to an organ involving
the wrong patient, wrong mode of treatment or a leaking source; (c) to a tissue or organ other
than the treatment site that exceeds by 50 cSv and 50%, the dose expected from the treatment,
had the written directive been followed. Permanent seed migration is specifically exempted.
Since agreement states are obligated to adopt an adverse-event reporting rule that meets the
essential objectives of 35.3045 by October 24, 2005, I will confine my remarks to the revised
reporting rule.
The ME threshold is intended to approximate levels of clinical relevance or at least clinical

concern. NRC's goal is to capture not only those errors with potential for patient injury, but also
those that suggest that the Licensee's radiation safety or technical QA programs are deficient 3 (p.
20330). NRC uses ME incidence both to assess the effectiveness of its overall nuclear materials
regulatory program4 and as a key performance endpoint in its new "risk informed, performance-
based" regulatory paradigm. Essentially, MEs are used to identify both individual Licensees and
generic treatment situations that need more prescriptive regulatory attention. 3
If the ME reporting program is to identify flawed QA and safety programs, its definition must be
(a) decidable, i.e., provide criteria for unambiguously distinguishing between medical events and
nonevents; and (b) relevant, i.e., identify only significant "QA failures" warranting review and
possible modification of the associated QA program.
Determining whether a treatment is an ME under conditions such as the wrong patient or

modality is straightforward. A dose delivery error in excess of 20% is also straightforwardly
decidable. Most of the 23 MEs reported in NRC's Nuclear Material Event Database (NMED) 3 for
FY 2002 involved incorrect source strengths, unit conversions, or decay corrections; inaccurate
reference point localization; dwell times from the wrong catheter; gross source-train
mispositioning; or intravascular brachytherapy device failure. NRC's Medical Policy Statement 5
effectively limits patient safety oversight to assuring that the physician's written directive is
followed. For ME decisions, the difference between the delivered treatment and that intended by
the authorized user is to be evaluated based on the treatment site and dose-specification criteria
chosen by the radiation oncologist and documented in the written directive. So long as the
prescription quantity (e.g., D90 for prostate brachytherapy) is within 20%, local discrepancies
between the planned and delivered dose distributions exceeding 20%, caused by limited precision
of seed positioning, do not constitute medical events. Finally, dose-calculation or estimation
techniques that have uncertainties approaching 20% are irrelevant to ME classification, so long
as these procedures adhere to the appropriate standards of practice.
Do the ME criteria identify significant QA events? As current QA practice standards are based
upon a target dose-delivery accuracy of 5%,6 discovery of an avoidable 20% dose-delivery error
is clearly grounds for re-assessing the effectiveness of one's QA program. While many MEs may
not have immediate clinical consequences, there is ample evidence that 15%–20% errors do
352
measurably alter clinical outcome in many settings.7,8 Previously, I have argued6 that the ME
definition significantly improves upon "wrong-site" misadministration by specifying a dose
threshold (at least 50 cGy and a 50% dose-delivery error) which eliminates many meaningless
events. Records of past MEs are perhaps the best measure of ME relevance. In this spirit, I
challenge my opponent to identify just one FY 2002 NMED ME that is undeserving of
investigation as a potential QA program flaw.
Rebuttal
In his first point, Dr. Amols attacks misadministration-reporting levels as arbitrary because they
may not lead to unfavorable changes in clinical outcome. The principal regulatory role of the ME
reporting rule is to identify licensees with deficient safety programs, with the assumption that the
best defense against avoidable patient injuries is a functional QA program. The patient reporting
requirement is a secondary consequence: once an error has been reported to the government that
may have health consequences, ethically and legally this information must be shared with the
patient. I agree that treating ME as a potential patient harm index undermines its value as a QA
performance index, and makes ME reporting unnecessarily disruptive of the patient-physician
relationship. Dr. Amols' assertion is probably true that geometric uncertainty in treatment
delivery may make 20% or even 50% errors to small tissue volumes a routine consequence of
radiation therapy. However, ME assessment is based upon compliance with the written directive,
and not assurance that the dose at every point in a complex 3D dose distribution is delivered
within 20%.
Dr. Amols' second point is that ME fails the decidability criterion. Some of his concerns are
specific to the about-to-be obsolete misadministration concept. "Treatment site" and "prescribed
dose" are defined not by the regulations but by the planning and delivery team themselves. For
prostate brachytherapy, these may be "prostate with specified margins" and "D90." In addition,
the authorized user has the option to revise the written directive in light of source positioning
errors at any point prior to treatment completion. The "wrong site" ME criterion could be
problematic, however. A single seed unintentionally placed outside the prostate capsule could
deliver a dose in excess of 50 cGy and 50% of the planned distribution to some non-target tissue.
My hope is that the NRC will temper its decisions with a modicum of clinical judgment, and
accept the authorized user's signature on the post-implant plan as indicating clinical acceptance
of these unavoidable deviations.
Dr. Amols' final point is that mandated corrective actions following an ME are meaningless, and
that errors must be accepted as inevitable. This view is clearly contrary to prevailing standards of
care, both as practiced and as codified in AAPM's Task Group reports.9 Clearly, anyone has a
finite probability of "punching in the wrong numbers:" the challenge is to develop a system of
redundant checks that minimizes the probability such errors will go undetected. TG-40 (p. 585)9
clearly states that errors in excess of specified thresholds should be viewed as failures of the QA
system.
REFERENCES
1. NRC, Code of Federal Regulations, 10 CFR Part 0-50, Washington, DC; U.S.
Government Printing Office (1993).
2. H. I. Amols and M. Zaider, "On the question of defining misadministration in
brachytherapy," Brachytherapy 1, 123–125 (2003).
353
3. "10 CFR Parts 20, 32, and 35: Medical Use of Byproduct Material," U.S. Nuclear
Regulatory Commission, Washington, D.C., Federal Register 67: pp. 20250–20397, 24 April
2002.
4. Annual Report to the Commission on Performance in the Materials and Waste Arena:
SECY-03-0060, April 18, 2003, U.S. Nuclear Regulatory Commission,
http://www.nrc.gov/reading-rm/doc-collections/commission/secys/2003/secy2003-0060/2003-
0060scy.html.
5. "Medical Uses of Byproduct: Material Policy Statement, Revision," U.S. Nuclear
Regulatory Commission, Washington, D.C., Federal Register 65: pp. 47654–47660, 3 August
2000.
6. J. F. Williamson, "On the question of defining misadministration in brachytherapy,"
Brachytherapy 1, 121–123 (2002).
7. L. Potters et al., "A comprehensive review of CT-based dosimetry parameters and
biochemical control in patients treated with permanent prostate brachytherapy," Int. J. Radiat.
Oncol., Biol., Phys. 50(3), 605–614 (2001).
8. R. G. Stock et al., "A dose-response study for I-125 prostate implants," Int. J. Radiat.
Oncol., Biol., Phys. 41, 101–108 (1998).
9. G. J. Kutcher et al., "Comprehensive QA for radiation oncology: report of AAPM
Radiation Therapy Committee Task Group 40," Med. Phys. 21(4), 581–618 (1994).
354
8.16. Medical technologies should pass performance and cost-

effectiveness review in Centers of Excellence before being released
for diffusion in the clinical community
Charles A. Kelsey and Gerald Cohen
Reproduced from Medical Physics, Vol. Volume 25, No. 7, p. 1099, July, 1998
OVERVIEW
In 1987 W. Schwartz stated in an article in JAMA: ‗‗Long term control of the rate of increase in
(healthcare) expenditures thus requires that we curb the development and diffusion of clinically
useful technology.‘‘ More recently, the National Cancer Institute announced (NIH GUIDE 26,
No. 28, August 22, 1997) a new initiative to establish a single national network of investigators
to perform multi-institutional clinical trials in diagnostic imaging related to cancer. The goal of
the initiative is ‗‗the creation of a Network that will serve as an instrument for the expert clinical
evaluation of discoveries and technological innovations relevant to imaging. Its activities will
result in the following: (1) the expeditious, reliable, and comprehensive clinical evaluation of
new imaging modalities; (2) the facilitation of technology development and translational
research relating to imaging in academia and in industry; and (3) development of improved
clinical trials methodology specifically related to imaging and the early detection of cancer.
Although the NCI initiative may or may not be intended to address Dr. Schwartz‘s concern, it is
conceivable that it could be used for that purpose by regulators and payers of healthcare services.
This is one of several controversies surrounding the NCI initiative. This edition of
Point/Counterpoint addresses these controversies.
Arguing for the proposition is Charles A. Kelsey, Ph.D. Professor Kelsey received his Ph.D. in
Nuclear Physics in 1962 from the Notre Dame University. In 1965 he joined the University of
Wisconsin Radiology Department. In 1975 he accepted a position with the University of New
Mexico as Chief of Biomedical Physics and Professor of Radiology to work on the Los Alamos
Pion Cancer Therapy Project. Since 1985 he has been working in diagnostic radiology and
radiation safety. He has served on the National Cancer Institute‘s Radiation Study Section and
numerous other government advisory agencies. He is intimately familiar with government
regulatory actions and procedures.
Arguing against the proposition is Gerald Cohen, Ph.D. Dr. Cohen was born in New York City,
earning his B.A. degree at Queens College and his Ph.D. in Physics at Purdue University.
Feeling the attraction of Medical Physics, he trained for a year with Dr. Kereiakes at the
University of Cincinnati Medical Center. The advent of the CT scanner started his career in
diagnostic imaging at Allegheny General Hospital in Pittsburgh in 1975. He moved to the
University of Texas Medical School-Houston in 1978, publishing over 50 papers and book
chapters on various aspects of diagnostic equipment and image evaluation. Dr. Cohen has been at
General Electric Medical Systems since 1984 in various roles involved with advanced
applications and technology development in diagnostic imaging.
For the Proposition: Charles A. Kelsey

355
Opening Statement
I write in favor of the proposition for two major reasons. First, it will help to base decisions
about the acquisition of new medical equipment on a sound scientific basis rather than on a
hodge-podge of individual preferences. Second, it will reduce health care costs.
Centers of Excellence will be able to develop performance criteria and evaluate the cost-
effectiveness of new technologies in a much more scientifically exact manner than can any one
user or combination of users. Centers of Excellence will provide a firm scientific foundation for
the clinical applications of new technologies. If Centers of Excellence had been in existence in
the early 1970s they would have undoubtedly shortened the path that CT technology followed
into the clinical arena. Even today we have few strictly scientific studies which show that costly
CT examinations are superior to more conventional examinations in the sense that patients
experience better outcomes with CT. I only need mention spiral CT as another example of an
unproven, unevaluated, technology, even though most clinicians ‗‗feel‘‘ or ‗‗believe‘‘ that it is
immensely valuable.
The use of Centers of Excellence to evaluate technologies before they are allowed to be
purchased by the end-user will reduce medical costs by eliminating the not-so-good products.
The current system of allowing the ultimate user to decide whether or not to adopt a new
technology based on anecdotal evidence is clearly inefficient and costly. Centers of Excellence
will reduce costs to the government and managed care organizations because the Centers will
perform the scientific experiments and evaluations needed to establish purchasing criteria and
regulations. Third party payers, managed care organizations and regulatory agencies will save
the money they otherwise would have had to spend to develop these data. Saving money is good;
therefore, Centers of Excellence are good.
There are those, misguided but of good heart, who may disagree with my point of view. They
may argue that large organizations are not efficient in making good choices. As examples, they
may point to the response of the big three automakers to the introduction of smaller, more
efficient cars; to the response of IBM to the introduction of personal computers; or to Kodak‘s
decision not to purchase the Xerox technology. They may do this to disprove my point. These
examples, however, actually prove my point, because even in the face of less than desirable
decisions, these companies have survived. The combination of Centers of Excellence, watchful
government agencies, and frugal third-party payers will help ensure that technologies are
allowed to diffuse into clinical medicine in a thoughtful, cost-effective manner.
Those of us who believe there is substance to Dr. Schwartz‘s words: ‗‗...we must curb the
development and diffusion of clinically useful technology‘‘ are favorably disposed to the
establishment of Centers of Excellence.
Rebuttal
Dr. Cohen comes out strongly for reducing cost through cost effectiveness studies and raises
some valid questions regarding performance evaluation of new technologies. He wants to know
which procedures should be used in assessing a new technology, when in the technology‘s
development to begin the assessment, what are the real costs of a technology and who will pay
for the assessment. These are legitimate concerns, but ignore the question of whether such
assessment should be required before the technology is released for diffusion into the medical
community. Is it better to allow the present laissez faire free market system to determine which
356
medical technologies survive, or have a scientifically sound evaluation system in place to ensure
that only technologies that have been reviewed for performance and cost-effectiveness are
allowed into the medical community? I believe it is better to have a firm scientific foundation
justifying a technology before its introduction. This foundation can best be established by centers
of excellence established solely for this purpose. History shows that once government entities
make a decision or establish a program they can be ruthlessly efficient in working toward that
goal. And efficiency is good. Large entities such as the Post Office, the Department of Defense,
and the Internal Revenue Service all serve as examples of what we can look forward to in the
coming years if these Centers of Excellence are established. Any and all small steps toward an
improved technology will be thoroughly studied and their improvement verified before being
introduced. We will all rest easier knowing the Centers of Excellence are watching out for us.
Against the Proposition: Gerald Cohen
Opening Statement
It is clear that the revolutionary advances in medical imaging since the invention of the CT
scanner in the early 1970s have, until recent years, occurred against a backdrop of rapidly rising
healthcare costs. The high cost of some medical imaging technologies, such as CT and MRI
scanners, have made them highly visible cost-increasing targets. As a result, there have been
voices raised calling for prior review and testing of the ‗‗performance‘‘ and ‗‗cost-effectiveness‘‘
of such systems before diffusion in the medical community.
Such an approach raises many thorny questions and issues, such as:
• What are the appropriate clinical procedures to assess on new imaging technologies? In most
cases, these can only be developed by the medical community through clinical development in a
variety of clinical environments. Could anyone have predicted 25 years ago that the CT scanner
would become the workhorse for diagnostic imaging of the body? Prior restraint on diffusion
would become a classic ‗‗chicken and egg‘‘ situation, with the result that the most potentially
clinically useful and cost-effective procedures might never see the light of day. Experience has
shown that the marketplace is smarter than the experts in predicting the future.
• At what stage in development should assessment be performed? Medical technology, unlike

drugs, is a continuously evolving and dynamic process, using constant feedback from clinical
users for development. Too early or limited an evaluation would hinder innovation and inhibit
useful technologies. In addition, if each new model or upgrade needed evaluation, this would
make improvements cost-prohibitive.
• What do we assume for ‗‗cost‘‘? Assumptions regarding the clinical environment, allocation of
overhead costs, system throughput, system price, etc. will greatly impact the results obtained.
‗‗Charge‘‘ is often used as a surrogate, but this is often an artificial number for billing purposes.
Do we assume the performance and throughput of a first generation system, or do we assume
future developments? Do we use the ‗‗costs‘‘ at a teaching medical center, a community hospital,
or an outpatient imaging center?
• Who pays for the trials? These potentially significant added costs have the potential for making
imaging technology less, rather than more, cost-effective.
357
Physicians and providers have responded to the healthcare systems‘ cost-sensitive environment
by demanding higher patient throughput, expanded versatility, greater quality, and decreased unit
costs, and manufacturers have responded accordingly. Physicians have judged medical imaging
to play a valued, cost-effective role in patient management, as reflected by its growing utilization
over the past four years, at the height of healthcare cost sensitivity. Processes such as
appropriateness criteria, practice guidelines and disease management should be supported and
encouraged to help guide physicians in the best use of imaging for the care of the patient.
Increased productivity should be sought everywhere. Cost-effectiveness studies should be
supported and encouraged, but should not become a regulatory barrier to diffusion.
Rebuttal
I think that Dr. Kelsey‘s final remark captures the essence of the debate. If one agrees with the
remarks attributed by Dr. Kelsey to Dr. Schwartz, that ‗‗...we must curb the development and
diffusion of clinically useful technology,‘‘ then I agree that the establishment of the Centers of
Excellence, as envisioned by Dr. Kelsey, can be a useful element in achieving that goal.
However, if one feels that the development of clinically useful technology is an important
element in improving the quality of healthcare, then perhaps we should examine alternative
approaches.
Centers of Excellence can be a useful means for developing an infrastructure capable of

performing selected clinical trials, technology assessments, and outcome studies, in a scientific
and efficient manner. Such research can serve as a useful guide for physicians, equipment
purchasers, and third-party payers in their purchasing and reimbursement decisions, and should
be supported. However, if such studies become a regulatory requirement before new technology
is ‗‗allowed‘‘ to be offered to the clinical community, then the supporters of curbing the
development and diffusion of new technology will have achieved their goal. Requiring a limited
number of Centers of Excellence to ‗‗approve‘‘ all new technology will severely narrow the base
of scientific and clinical research, slow the rate of technology development, and significantly
increase the cost of development.
358
CHAPTER 9
Education
9.1. Medical physics graduate programs should focus on education

and research and leave clinical training to residencies
Gary T. Barnes and Timothy K. Johnson

Reproduced from Medical Physics, Vol. 26, No. 10, pp. 2051-2053, October, 1999
OVERVIEW
Traditionally, medical physics predoctoral and postdoctoral educational programs have provided
a mixture of didactic instruction and on-the-job training in a clinical environment. Graduates of
these programs have taken jobs where usually, but not always, their work is supervised by an
experienced medical physicist. Many, perhaps most, of the leaders in medical physics have
followed this pathway into the field. Residencies in medical physics are a relatively recent
innovation. With their advent, the Commission on Accreditation of Medical Physics Education
Programs, Inc. (CAMPEP) has chosen to emphasize didactic training in medical physics
educational programs. Programs that place too much emphasis on clinical training, and not
enough emphasis on didactic training, are not granted accreditation. CAMPEP‘s position has
jeopardized the future of some medical physics educational programs that have produced
credible graduates over several years. Whether this position reflects the best interests of the
medical physics profession is the subject of this Point/Counterpoint issue.
Arguing for the Proposition is Gary T. Barnes, Ph.D., Professor and Director of the Physics and
Engineering Division of the Department of Radiology at the University of Alabama at
Birmingham Hospital and Clinics. Dr. Barnes is a past president of the AAPM, served on the
AAPM Commission on Accreditation of Educational Programs for Medical Physicists for five
years (1991– 1995). He assisted in transition of the AAPM Commission‘s responsibilities to the
current independent Commission, and in 1996 chaired the CAMPEP Graduate Education Review
Committee.
Arguing against the Proposition is Timothy K. Johnson, Ph.D. Dr. Johnson is Associate
Professor in the Department of Radiology, and Director of the Graduate Program in Medical
Physics at the University of Colorado Health Sciences Center. He is the author of the
MABDOSE internal radionuclide dosimetry software [see Medical Physics 26, 1389–1403
359
(1999)], and has high hopes for radioimmunotherapy as a supplement to external beam and
chemotherapy in the treatment of cancer.
FOR THE PROPOSITION: Gary T. Barnes, Ph.D.
Opening Statement
The careers that medical physicists embark on are many. The majority involve the clinical
support of radiologists, radiation oncologists, and other medical disciplines. Many work in large
medical center departments and have additional responsibilities involving postgraduate physician
training, research, and/or medical physics graduate education. A smaller percentage work in
industry in the areas of product development, support, sales or marketing. A requirement of all
areas is a sound didactic foundation. This can only be achieved with extensive and
comprehensive formal course work. Such a foundation is a prerequisite to cognitive problem
solving and facilitates intellectual growth following matriculation. In addition, in the hospital
environment one must have a good understanding of medical physics clinical responsibilities.
These responsibilities take time to assimilate to the level that they can be performed in a
professional manner without supervision.
Medical imaging and radiation oncology have become more complex in recent years as have
medical physics and its subdisciplines. In 1960 imaging consisted mostly of radiography and
conventional fluoroscopy. In larger departments nuclear medicine imaging was done with
rectilinear scanners.
Radiation oncology was limited for the most part to cobalt-60 teletherapy, orthovoltage units and
implants. The picture in 1999 is much different. Imaging utilizes radiography, fluoroscopy,
planar gamma camera images as well as CT, ultrasound, MRI, SPECT and PET. Highly complex
digital heart catheterization and interventional radiology laboratories are commonplace. Medical
centers are making significant investments in digital radiography and PACS. Dual modality
linear accelerators, high dose rate brachytherapy, CT simulators, computerized 3D treatment
planning, and conformal radiation therapy are commonplace in radiation oncology. Intensity
beam modulation and stereotactic radiation therapy are being utilized at larger medical centers.
The knowledge and experience that a clinical medical physicist must have to function as a
professional is much greater today than it was in the 1960s. This knowledge and experience
cannot be achieved with two years of graduate school and the writing of a thesis or even with
additional years devoted to research and a Ph.D. dissertation. It requires additional hands-on
working experience under the supervision of experienced medical physicists. This fact is readily
apparent to the certifying boards which require three or more years of working experience before
one can sit for an exam in one of the subdisciplines of medical physics. To date recent graduates
have obtained this experience by either taking a junior level position or by entering a medical
physics residency. The advantage of the latter is that it provides a broad level of structured
working experience under a group of medical physicists. Claiming that a two year graduate
program can provide this level of experience, and that at the end of the program the individual is
a professional and can work independently and without supervision, is not realistic. Based
on the experience requirements of the American Board of Radiology and the American Board of
Medical Physics, it is misleading.
360
In summary, for an individual to be a clinical medical physicist, both a good didactic graduate
education and extensive hands-on experience are required. Both requirements are recognized by
the American Board of Radiology and the Board of Medical Physics. The time it takes to achieve
this level of competence should be formally recognized and is best accomplished by a structured
medical physics residency following the completion of graduate school. Salaries in medically-
related professions are related to supply and demand, and also to the time it takes to obtain the
necessary hands-on experience. In the medical field working experience is achieved initially and
formally in structured residencies. This model works well, is accepted by the medical field and
should be employed in clinical medical physics.
Rebuttal
Dr. Johnson and I are in agreement on several points. We agree that it is important for a medical
physicist to have a sound didactic background. We also agree that professional skills are
necessary for a medical physicist to function successfully (and independently) in the clinical
environment, and that these skills are not learned in the classroom. A difference in our positions
is that he views the Master of Science degree in medical physics as a professional degree,
whereas I view it as an academic degree. A more fundamental difference in our positions is that
he argues that a sound didactic foundation and satisfactory level of professional skill can be
obtained in a master‘s degree program. It is my position that this cannot be accomplished in the
time typically required to obtain a master‘s degree (i.e., one and one half or at most two years).
The recommended course work outline in AAPM Report No. 44 ‗‗Academic Program for Master
of Science Degree in Medical Physics‘‘ is extensive and will completely occupy a student‘s time
unless it is watered down and taught without appropriate rigor and mathematical sophistication.
Furthermore, AAPM Report No. 44 does not mention computer courses or a formal course in the
mathematical methods of medical physics. I consider these courses to be essential to permit
students to adapt to the current clinical and alternative (i.e., research, product development,
etc.) job climate, and to future advances in the field.
Dr. Johnson argues that several medical physics programs incorporate the development of
clinical skills into their curricula. He believes it would penalize graduates of these programs to
subject them to additional years of indentured servitude, and that this procedure is just plain
wrong. However, the course work requirements fully occupy the student‘s attention for the time
typically required to obtain a master‘s degree. Adding the development of a significant level of
clinical skill into the curriculum can only be accomplished in one of two ways: either water
down the course work or extend the degree time. Although watered-down courses are easier to
teach, they short-change students and limit their future growth. Dragging out the time for a
student to obtain a degree is, in my opinion, extending the student‘s indentured servitude under
far less satisfactory conditions than spending time in a structured residency. In a medical physics
residency the pay and fringe benefits are the same as physician residencies, and are significantly
better than the pay and benefits of a graduate student.
AGAINST THE PROPOSITION: Timothy Karl Johnson, Ph.D.

Opening Statement
Medical Physics is a professional degree based on an applied science. When individuals train for
the field, it is with the reasonable expectation that they will be employed in a clinical setting
possessing certain skills. This is, after all, where the jobs are: A quick scan of the AAPM‘s
monthly Placement Service ‗‗blue book‘‘ demonstrates that the majority of employment
361
opportunities are in Radiation Oncology, and that the professional skills desired are (1) Machine
and source calibration; (2) treatment planning for external beam and brachytherapy sources; and
(3) checking treatment charts. These skills are not learned in the classroom.
While a background knowledge obtained through didactic course work is necessary for a broad-
based understanding of radiation, it is hardly sufficient for the majority of tasks that evidently are
expected from the degree. This being the case, it appears irresponsible to provide a program
where didactic lectures are emphasized at the expense of the very skills that make one
employable.
Reserving clinical training to the residency appears to be motivated by two considerations: (1) an
effort to place Medical Physics education within the same framework as the Radiology residency
for reasons of professional stature and parity; and (2) a lack of uniformity in Graduate program
curricula that prevents a standards committee like CAMPEP from avoiding subjective
evaluations. The medical school curriculum, however, has evolved to a natural division between
basic skills learned in the first four years, and specialized clinical skills imparted in a residency.
In contrast, a number of medical physics programs incorporate the development of clinical skills
into their curricula. To penalize graduates of these programs to additional years of indentured
servitude is just plain wrong.
With respect to the second point, a certain amount of diversity exists among medical physics
graduate programs. Diversity always plays havoc with standards, because an unintentional
outcome of standards is often homogenization. Standards make evaluation easier on the part of
examiners charged with overseeing individual program compliance; they do not allow for
programs that supercede the standard‘s baseline to be recognized for any excellence that goes
above and beyond the published standard.
One needs to ascertain whether there are alternative means for distinguishing graduate programs
in Medical Physics. A more natural metric for distinguishing ‗‗ability‘ might be a two-tier
accreditation. The two-tier evaluation would allow CAMPEP to categorize the major differences
in graduate medical physics programs. Graduates having a clinical component would have a
different level of competence associated with their degree. Depending on a student applicant‘s
professional goals and aspirations, they could opt for a clinical residency depending on the tier
their graduate school prepared them for. When the clinical training occurs should not matter:
only that it occurs.
Rebuttal
Dr. Barnes reiterates the general argument for a sound foundation in didactic course work,
namely that it facilitates general problem solving skills and fosters intellectual growth. This is
followed by statements regarding the value of clinical experience. Neither are in dispute.
However, he goes on to state that the knowledge and experience of a clinical medical physicist
cannot be achieved within a typical two-year medical physics graduate school program, and that
the advantage of the clinical residency is that it provides a broad level of structured work
experience under a group of medical physicists. This sounds very much like the advantage of our
graduate program, in which the hands-on experience is taught in concert with related didactic
learning. I encourage prospective students that are reading this to apply
to our program.
362
Dr. Barnes states that in the medical field, working experience is achieved initially and formally
in structured residencies. For physicists, the initial work experience is typically gained within the
framework of a graduate program, followed by supervised positions in the field. The need to
separate the work experience from didactic lectures is not the defining model of residency
programs. In fact, radiology residents are usually exposed to a core lecture series that includes a
didactic physics course concurrent with their clinical experience.
Upon talking to the few Colorado graduates that chose to complete a clinical residency, the
‗‗new skills‘‘ that they ‗‗mastered‘‘ to a large extent centered on learning to use a different
interface for a treatment planning computer. This was not substantive new knowledge, but rather
the consequence of their taking a job that employed a treatment planning system different from
the one they trained on. The physics of ionizing radiation interacting with matter did not
change—just the appearance of how to manipulate it. Moreover, these skills were gained within
several weeks, not one year of concentrated effort. An unbiased observer would probably
conclude that this was the period of time that any medical physicist (board certified or not)
would require to master a new dosimetry system.
I do not think anyone believes that a two-year graduate program provides a level of experience
whereby a graduate can work independently and without supervision. It does provide a level of
experience whereby a graduate should be able to command an entry-level professional salary
without being subjected to the penury associated with student status.
363
9.2. Physicists directing training programs should cooperate to align

student enrollment with market demand for medical physicists
Don Tolbert and James A. Zagzebski

Reproduced from Medical Physics, Vol. 27, No. 7, pp. 1487-1489, July, 2000
OVERVIEW
Many persons believe that the market demand for medical physicists is bound to diminish as a
result of growing fiscal constraints imposed by the increasingly competitive environment of
managed care. Some would say that the market demand for physicists has already begun to sink.
They might argue that it is unwise for the discipline and unethical for the training programs to
graduate more medical physicists than the market can absorb. They would endorse an agreement
among training program directors to limit student enrollment so that it is aligned appropriately
with the market demand for physicists. Opponents to this point of view might suggest that
market demand is impossible to predict, and that one should be optimistic that the growing
sophistication of diagnostic imaging and radiation oncology will require more medical physicists
in the future. They also might propose that an academic discipline such as medical physics has
no right to deprive interested young people from pursuing studies in the field, and that no
discipline guarantees employment to its graduates. This debate is the subject of this month‘s
Point/Counterpoint issue.
Arguing for the Proposition is Don Tolbert, Ph.D, FAAPM, FACR. Dr. Tolbert earned his Ph.D.
in Nuclear Physics at the University of Kansas in 1968. Following two years as a Post Doctoral
Fellow at Florida State University, he accepted a Post Doc Re-Trainee position with the Medical
Physics Division of Radiology at the University of Wisconsin in Madison. After seven years on
the faculty he moved to Hawaii where he became Director of a Radiological Physics Outreach
program. From 1981 through 1990 he was Managing Partner of Mid-Pacific Medical Physics and
since 1991 has worked at Tripler Army Medical Center in Honolulu, HI.
Arguing against the Proposition is James A. Zagzebski, Ph.D. Dr. Zagzebski is chairperson of
the Medical Physics Department of the University of Wisconsin. He teaches courses and
conducts research in diagnostic ultrasound physics. Currently he and his students are working on
tissue feature extraction with ultrasound, flow measurements using ultrasound contrast agents
and problems in 3-dimensional imaging. Dr. Zagzebski received his Ph.D. in Radiological
Sciences in 1972 and has been with the University of Wisconsin since that time.
FOR THE PROPOSITION: Don Tolbert, Ph.D., FAAPM, FACR
Opening Statement
I write in favor of the proposition. Two conditions are necessary however. The planned objective
must benefit (a) both the state and federal government, and (b) the medical physics profession.
Item (a) is necessary to avoid antitrust litigation. We should petition federal and state authorities
having jurisdiction over medical physics to help arrive at the appropriate number of program
364
positions to ensure an adequate supply of graduates while conserving increasingly scarce public
resources.
How does this benefit medical physics? Talent and experience from basic science pursuits have
traditionally been important to the profession. Our founding fathers began not only the research
and development aspects of medical physics, but also the clinical applications. Medical physics
training programs accredited by the Commission on Accreditation of Medical Physics Education
Programs (CAMPEP) Are now the preferred choice of most students pursuing a medical physics
career (Bhudatt Paliwal, Chairman of CAMPEP, personal communication). While research and
development aspects of medical physics may continue to benefit from individuals entering
directly from the basic sciences, the clinical aspect of the profession does not.
In all aspects of our profession the standard has been raised considerably. Employers now expect
medical physics graduates to function at a level comparable to physicians and other
professionals. The influence of managed care however, has enabled economic factors to
outweigh professional competence in hiring. When the output from medical physics training
programs does not meet market demand, entry into the field occurs through the side door where
the benefit of medical physics training is not structured per CAMPEP accreditation criteria.
Entry through the side door disadvantages the medical physics profession and exposes
institutions to greater liability.
Program directors must make decisions based on accurate numbers representing dynamic
variables in the marketplace. Much information could come from the membership data gathered
and maintained by the AAPM Headquarters. For example, if the AAPM membership in the
categories of Canadian, Student, and Honorary are subtracted from those for the Total
membership, a linear regression analysis shows an increase of approximately 126 members per
year (Sal Trofi, Jr., AAPM Executive Director, personal communication). Does this increase
represent market demand? CAMPEP accredited training programs graduate an average of
approximately 100 per year (Bhudatt Paliwal, personal communication). If the market demands
126 per year and the average accumulated total from accredited training programs is 100 per
year, then non-accredited programs must be making up the difference, or, the difference must be
entering through the side door.
Rebuttal
I agree that excellence must continue to be a priority in our training programs. This has provided
much leadership in medical radiological applications and is certain to play an increasingly
important role in non-radiological efforts.
Cooperation by program directors would have to include consideration of the dual nature of
training programs. At the risk of oversimplification, some graduates are employed in research
and development while others are employed in the clinic. Jim states that the medical physics job
market is ‗‗ . . . a market whose demands have a time constant that seems shorter than the
training period for students.‘‘ An examination of (a) student membership and (b) salaries for all
degrees (certified) with primary employment in radiation therapy1 suggests otherwise.
Student AAPM membership enrollment from 1990 through 1999 is distinctly bimodal. From
1990 through 1995 a linear trend line (R2=0.925) shows an increase of over 55 students per year,
while from 1995 through 1999 a linear trend line (R2=0.909) shows a decrease of over 25
students per year. The same analysis applied to average ‗‗Primary Income‘‘ salaries for all
365
degrees (certified) employed in radiation therapy show similar results. From 1990 through 1994
a linear trend line (R2=0.976) indicates a salary increase of $4,710 per year while from 1994
through 1998 the same analysis (R2=0.998) indicates a salary increase of only $1,950 per year.
These numbers are only suggestive. The ‗‗student‘‘ membership category is subject to error
because it can take a few years to have the membership category accurately reflect passage from
training to non-training status. The ‗‗all degrees‘‘ category was used only because it was
continuously used from 1990 through 1998. The numbers above suggest that the market trend
may remain constant for a period of four to five years. I believe this is long enough to allow
program sizes to adjust to market demands.
A model should be developed that uses membership, salary survey, etc., information as input
data for suggesting when market cycles may change.
AGAINST THE PROPOSITION: James A.: Zagzebski, Ph.D.
Opening Statement
Directors of Medical Physics training programs certainly owe it to the profession to be attentive
to the medical physics job market as they do long range planning. However, they should not
attempt to fine tune program enrollments to respond to ups and downs of the medical physics job
market, a market whose demands have a time constant that seems shorter than the training period
for students. Instead, directors should focus on maintaining high quality programs, programs that
are responsive to the needs of the field, and most importantly, programs that foster discovery.
Five years ago a graduate of our program approached me and expressed dismay that medical
physics training programs had not reduced their output of graduates. It seemed the ‗‗flood‘‘ of
new medical physicists into the job market at the time, along with pressures on the health care
dollar, were reducing opportunities for established physicists. Luckily, few programs cut
enrollments at that time. Had they done so, employers in today‘s market would be even more
pressed to find well-trained physicists than they currently are.
While market demands eventually will influence the size and breadth of all training programs,
directors should work for excellence in these programs rather than try to outguess job market
pressures. Men and women choose medical physics as a career because they enjoy the ‗‗physics
approach‘‘ to discovery, and they have a keen interest in leaving this world in slightly better
shape than it was when they entered. The physics profession, the medical physics community,
and society, has benefited greatly from the thousands of fine individuals who have been trained
as medical physicists. Who doubts that their contributions will continue?
If we impose constraints on the size of medical physics programs, then, in a sort of self-fulfilling
prophecy, we also will be constraining growth in our field. Some of the most important
discoveries in medical physics have been fostered within our training programs, with graduate
students working alongside faculty. Technology such as Cobalt 60 machines, computerized
treatment planning, DSA, digital imaging, bone densitometry, SPECT, PET, TLD, CT, MRI, and
3-D ultrasound have emerged from our academic programs. Today their atmosphere of discovery
is spawning MRI angiography, biomagnetism, tomotherapy, thermal therapies, and molecular
imaging. It is important to foster this
366
atmosphere—which is creating the jobs for tomorrow‘s medical physicists—rather than thwart it.
Great things happen when medical physics programs expand their enterprise, and along the way
create the technology that will be needed in tomorrow‘s medical practice.
Rebuttal
Don is correct in noting that CAMPEP accredited programs are the preferred entry into our
profession. The premise, however, that the number of graduates can be coupled to the medical
physics job market ignores key factors in training medical physicists. Interestingly, these factors
also are linked to the high professional standards required of medical physicists, also noted by
Don.
The first factor is that there is keen competition among graduate programs in attracting well-
qualified applicants. Most medical physics programs compete not only with each other but also
with graduate physics programs for outstanding undergraduate physics majors. If only a handful
of qualified individuals applies to a program, should admissions standards be lowered to fill
professionally-based quotas? Of course not! Vice-versa, if a program is blessed with an
unusually large number of highly qualified applicants, and faculty have means of supporting
these students, shouldn‘t these young physics majors be given the same opportunities to
contribute to the field that we have enjoyed? This might not occur with mandated enrollment
limitations.
The second factor is space limitations and limited training resources in graduate medical physics
programs. Programs have finite laboratory and equipment resources to provide training at the
quality required to assure success in today‘s complex environment. Ramping up training slots to
fill immediate job quotas would cause chaos within a training program, as this may require
additional faculty and additional equipment. Vice-versa, why would a program waste precious
training resources, and deny entry to a competent student when the job market happens to be
tight (assuming program directors are not misrepresenting the job market to their applicants)?
Manpower needs in medical physics must be dealt with at the professional level, while programs
should be allowed to respond to market demands as they see fit. Individual physicists can do
much to introduce qualified undergraduates to our field and to encourage them to apply to
CAMPEP training programs. It is surprising how many undergraduates are still not aware of
medical physics as a career option. Professional organizations such as AAPM can continue to
support training with fellowships, reduced registration fees for students, etc. When we enjoy an
overabundance of graduates, these activities can be curtailed. But medical physics training
programs should concentrate on maintaining high quality programs, enrolling the number of
students they can support, and continuing to provide outstanding training experiences.
REFERENCE
1. AAPM Professional Information Survey Report.

367
9.3. All medical physicists entering the field should have a specific
course on Research and Practice Ethics in their educational
background
David Switzer and Nicholas Detorie

OVERVIEW
There can be little disagreement with the premise that medical physicists must function ethically
in all endeavors, including research, practice and education. How ethical principles are best
instilled into medical physicists can be argued, however. Some would say that every physicist
should take a specific course on ethical principles and their applications. Others believe that
ethical behavior is best learned through association with ethical teachers, mentors and colleagues.
Still others might argue that ethical behavior is culturally-instilled, and a separate course in ethics
is not needed. How ethical principles are assimilated by medical physicists is the topic of this
Arguing for the proposition is David Switzer, MS. Mr. Switzer received his Masters Degree
(Major Physics, Minor Mathematics) from Saint Louis University, his thesis dealing with
thermionic emission using x-ray crystallography, high vacuum, high voltage, and electron
diffraction studies in completion of the research. He began his career in Medical Physics at the
University of Missouri in 1969. He has served as Chairman of the Ethics Committees of the
AAPM and the ACMP. His place of work for the past seventeen years has been the Northern
Rockies Cancer Center in Billings, Montana. Mr. Switzer is certified by the ABR (TDRP) and
ABMP (ROP) and is a Fellow of the ACMP.
Arguing against the proposition is Nicholas Detorie, Ph.D. Dr. Detorie is the Interim Director of
Medical Physics in the Department of Radiation Oncology at Johns Hopkins University and is the
Director of the Medical Physics Residency Program at JHU. Dr. Detorie has served as Chair of
Public Education for the AAPM and as Chair of Continuing Education for the ACMP. Currently,
he Chairs the Medical Physics Standards Committee for the ACR Commission on Physics and is
the AAPM representative to the AIP Media and Government Relations Committee. He is also a
member of the Editorial Board for the AAPM Newsletter.
FOR THE PROPOSITION: David Switzer, M.S.
Opening Statement
Medical physics emerged in the modern era as an outgrowth of innovations fueled by

technological advances made during World War II. These advances included the emergence of
microwave technology, the creation of artificial radioactive material, and the use of computers
powered by solid-state devices. In the late 1940s, the embryonic nucleus of our present field
came into being. Pioneering medical physicists were supported, encouraged, and challenged by
their physician colleagues to assist in diagnosis and treatment with the innovative use of new
368
technologies. Those physicists were well served by their strong scientific education, training, and
experience.
The preparatory training of medical physicists has traditionally not included ethics. I am not
alleging that medical physicists lack ethics. Ethical issues accompany many medical physics
applications, such as calculations to determine the radiation dose to critical organs or a fetus.
Medical physicists are well aware that the future holds other roles requiring ethical applications
of our knowledge.
Medical physics services often include the provision of advice on whether or not to provide
treatment, the use of curative versus palliative courses, the potential occurrence of early and late
effects from radiation treatments, and the loss of function or changed morphology related to the
disease and/or its treatment. Medical physicists know that physicians, nurses, radiation therapists,
administrators, and attorneys take formal courses in ethics. As physicists we could further our
professional standing if our credentials included ethical courses equal to those taken by our
colleagues. Training in ethics would expand the respect we have for each other and for our
practice as part of a professional team.
Ethical behavior is closely connected to our cultural background and heritage. It is honed by our
work with mentors and colleagues. Medical physicists could benefit by having an initial
introductory course in ethics, which would provide a uniform basis for establishing soundly-
based ethical positions. Medical physicists would then be better prepared for the many
professional dilemmas that inevitably arise. In a recent article, N. Reed Dunnick said, "It is not
enough that each of us individually tries his or her best to practice ethical behavior. We must
include it in residency training."1 This is also appropriate for medical physicists who have not
had an ethics course in their graduate education. The first guideline in the AAPM Ethics
Guidelines is "Medical physicists should strive continually to improve their knowledge and
skills . ."2 We should strive to improve our ethical decision-making skills.
Concerns which might be addressed in a formal ethics course would include: self-respect and
professional respect; collegial communications; professional respect for each other; pursuit of our
work as professionals; and ethical challenges in employment situations. 3 Ethical positions could
be developed on issues such as relations with patients and medical colleagues, allocation of
medical resources, and relations with vendors.4,5,6 A short course at the RSNA, AAPM, ACMP,
or ASTRO meetings could be held for those of us now in the field.
Rebuttal
During preparation of this paper, honesty, ethical behavior, and integrity have been part of the
business news. There are calls within the professional business community to teach ethics in
order to set standards and guides for improvement. For example, "Common features of an
effective ethics and awareness training program include: (1) live instruction, (2) small class sizes,
(3) significant group interaction, (4) at least 4 hours of training, and (5) separate courses for
compliance areas. In any organization, a good business ethics process is the first and most
important line of defense against unethical or illegal activities. Training also helps improve
their (employees) decision-making capabilities in the presence of ethical dilemmas as well as
establish goal congruence between employees and the organization."7 Societal expectations have
increased in recent years for all medical professionals. Medical Physicists' professional
colleagues have chosen to include ethics courses in their curricula. There is a need to meet the
public's expectations.
369
The Ethics Committee of the AAPM has dealt with issues of acknowledgement of corporate
support for research work, review of others' work, etc., and issued formal statements included in
the Directory.2 Disputes have occurred between Medical Physicists who could have been served
by a formal ethics course. Conflict, heartache, and animosity could have been averted.
In many Medical Physicist's careers there have been exceptional physicists, medical physicists,
physicians, and other colleagues who have provided significant insight, mentoring, and
leadership. In many cases these colleagues were innovators of a technique, method, or device.
These colleagues did not take a course in the topic for which they were the initial expert. Mother
Teresa is rightfully an example of an individual—a pioneer—who was sainted for her devotion to
the cause of the poor, the sick, and the hungry. May we all look to her as an example for ethical
behavior.
AGAINST THE PROPOSITION: Nicholas Detorie, Ph.D.
Opening Statement
Mother Teresa did not, to the best of our knowledge,8 have a specific course in ethics. Would
anyone doubt that she was an ethical person? This 20th century icon exhibited behavior based on
the fundamental building blocks of ethics: honesty and integrity. She specifically illustrates that a
course in ethics is not required for ethical behavior or even sainthood! We may not all be like
Mother Teresa, but her example brings into focus two important issues related to the requirement
of a specific educational "ethics" course: purpose and need.
From our professional perspective, what would be the purpose of such a course? Surely, it could
not instill honesty and integrity, since these attributes are not learned by short-term exposure to
discourse and dialogue in a traditional classroom setting. Honesty and integrity are assimilated
over the long-term by association with individuals that behave in like manner. The purpose of a
specific course would be acquisition of knowledge per se. Acquired knowledge in any profession
requires a priority process. Becoming a Medical Physicist requires curricula that develop
professional skills. Consequently, knowledge priorities are established emphasizing the technical
aspects that help us contribute to better patient care. With knowledge per se as purpose, the
ethics course yields little patient benefit. Does this mean that "ethical knowledge" is
unimportant? No. It means that ethics is not a training priority compared with other knowledge
that must be mastered in order to be professionally effective.
The second issue to consider is need. Think about various ethical issues that may arise in our
professional practice. Plagiarism, data falsification, conflict of interest, software piracy,
confidentiality, and author acknowledgement are just some that may come to mind. Have we not
demonstrated an ability to deal with these issues? Is it possible that as a professional group,
physicists are not behaving properly, and therefore need a course with a specific ethics focus? To
address this question a National Science Foundation Grant was awarded to Wylo and Thomsen to
conduct a survey of the physics community. They asked the question: Should physics students
take a course in ethics?—Physicists Respond.9 Survey results indicated that most physicists
thought such a course would be helpful, but approximately only one third of the physicists in any
given category thought an ethics course should be required. The perceived need for such a course
is not strong.
370
A simple test may reveal your own bias regarding this issue. After reviewing the resume of two
applicants for a medical physics position in your radiation oncology department—all other items
and issues being equal—do you want to pursue the candidate that has the additional course in: (a)
Physical Techniques for IMRT Planning and Delivery or (b) Research and Practice Ethics in
Medical Physics. Picking (a) indicates your implicit agreement that a specific ethics course is not
warranted at this time.
There is a role for ethics education in our profession that can be met by seminars and symposia
that may sensitize us to current ethical issues. However, there is no well-defined purpose or
documented need indicating that medical physicist entering the field should be required to take a
specific ethics course.
Rebuttal
Mr. Switzer's asserted purpose for a specific ethics course for medical physicists includes:
enhanced professional standing, uniform basis for ethical decision making, improved "respect"
for self and colleagues and improved ethical knowledge. He has tried to identify a need for a
specific course by enumeration of specific instances involving ethics: fetal dose calculations,
advice to physicians regarding treatment, probability of radiation effects, employment
relationships, and relationships with patients, medical colleagues, and vendors. His "ethics"
purpose identifies only peripheral priorities and his need lacks validity.
The Medical Physicist is to the Radiation Oncologist/Radiologist as the Anesthesiologist is to the

Surgeon. We improve patient care and aid our medical colleagues by providing professional
competence, judgment, and technical consultations hewn from physics training. Behaving this
way automatically enhances our professional standing amongst colleagues and patients and
develops relationships of "respect." This behavior cannot be learned in a specific course. Mr.
Switzer tacitly admits that ethical behavior is accrued over the long term via "background and
heritage." The additional purpose, to improve knowledge via AAPM guidelines, realistically
means Medical Physics knowledge, not knowledge of other subjects.
Regarding need, Mr. Switzer concurs that he is not claiming AAPM medical physicists lack
ethics. I support his claim. My inquiry with the past and current chairs of the Ethics Committee,
Dr. Hrejsa and Mr. Freedman,10 indicates the yearly number of ethical "cases" has remained flat
and may be decreasing. About 1 "case" per year potentially requires action. With almost 5000
AAPM members, the demonstrated need for a specific ethics course is practically nonexistent.
Without a better-defined purpose and need, medical physicists entering the field should not be
required to have a specific course in their background. We should not consider surgery when a
band-aid will do. Educational symposia are sufficient for our needs.
REFERENCES
1. N. R. Dunnick, "Ethics: Why it belongs in residency training," ACR Bulletin 59(2), 12–
13 (2003). First printed in ARRS Memo Fall 2002 issue.
2. AAPM Directory, "Guidelines for ethical practice for medical physics," pp. 50–52
(2003).
3. Introduction to the Professional Aspects of Medical Physics, 95, edited by K. R.
Hogstrom and J. L. Horton (Cowart, Ronald W.).
371
4. D. Stern, "Current activities in teaching professionalism and ethics," University of

Michigan Medical School, personal communication from N. Reed Dunnick.
5. Program in Medical Ethics, Indiana University School of Medicine, "Ethics Education for
Medical Students," http://shaw.medlib.iupui.edu/ethics/students.html
6. "The international code of medical ethics," World Medical Association Bulletin 1(3),
109–111 (1949), http://gwdg.de/~uelsner/internat.htm
7. L. Ponemon and A. Felo, "Key features of an effective ethics training program,"
Management Accounting 78(4), 66–72.
8. Private communication, "Missionaries of charity," Washington, D.C., March, 2003.
9. B. Wylo and M. I. Thomsen, Should Physics Students Take A Course In Ethics?—
Physicists Respond, Ethical Issues in Physics Workshop Proceedings, Workshop II (July, 1996),
http://www.physics.emich.edu/mthomsen/ethtaboc.htm
10. Private communications, May 2003.
372
9.4. Physicists are better educated for a career in medical physics if

they graduate from a specialized medical physics graduate program
rather than from a more traditional physics graduate program
Ervin B. Podgorsak and David W. O. Rogers
OVERVIEW
Persons entering the practice of medical physics today are more likely to have graduated from a
graduate program in medical physics than from a traditional physics graduate program. Some
might say this is a good trend because entering physicists have a greater knowledge of the
practice of medical physics. Others might say that to be a good medical physicist, one must first
be a good physicist, and a graduate degree in traditional physics helps ensure the latter. They
would claim that knowledge about the practice of medical physics can be acquired during
postgraduate residency training. These opposing points of view are debated in this month's Point
Counterpoint.
Arguing for the Proposition is Ervin B. Podgorsak, Ph.D. Dr. Podgorsak graduated in physics
from the University of Ljubljana in Slovenia in 1968 and then obtained his M.Sc. and Ph.D.
degrees in Physics from the University of Wisconsin in Madison. During 1973–74 he held a post-
doctoral fellowship at the Ontario Cancer Institute in Toronto. Since 1975 he has been employed
at McGill University in Montreal, where he currently holds positions of Professor of Medical
Physics, Director of the Medical Physics Unit in the Faculty of Medicine, and Director of the
Medical Physics Department at the McGill University Health Centre. He is board certified by the
CCPM and the ABMP.
Arguing against the Proposition is David W. O. Rogers, Ph.D. Dr. Rogers has just taken up the
Canada Research Chair in Medical Physics in the Physics Department of Carleton University. He
has been part of the graduate program in Medical Physics at Carleton since 1986. Previously he
worked at the National Research Council of Canada where he headed the Ionizing Radiation
Standards group since 1985. His research is centered around radiation dosimetry-related
measurement standards, clinical dosimetry protocols and the development and application of
Monte Carlo techniques to medical physics problems.
FOR THE PROPOSITION: Ervin Podgorsak, Ph.D.
Opening Statement
During the past two decades medical physics has undergone a tremendous evolution, progressing
from a branch of science on the fringes of physics into an important mainstream discipline that
can now be placed on an equal footing with other more traditional branches of physics. To be
productive in any of the traditional specialties of modern physics, physicists must not only
possess a solid knowledge of general physics and science, but also a rigorous didactic graduate
training in the specialty.
373
Many believe that medical physics is exempt from the requirement of didactic M.Sc. or Ph.D.
training in medical physics, and prefer a model in which physicists with a graduate degree in a
"straight" physics discipline can become a medical physicist through on-the-job academic and
clinical training in medical physics. For practical reasons, this approach was historically the
standard path to entering the medical physics profession. Now, however, this entry model should
be discouraged in favor of a model that provides a well-defined and rigorous four-step
progression to becoming a qualified medical physicist. The four steps are:
(1) Undergraduate degree in physics.
(2) Graduate degree in medical physics from a Commission on Accreditation of Medical Physics
Educational Programs (CAMPEP)-accredited program.
(3) Residency in one of the medical physics specialties (e.g., radiotherapy physics, diagnostic
radiology physics, etc.) at a CAMPEP-accredited institution.
(4) Certification in the particular medical physics specialty by an appropriate certification body
(e.g., American Board of Radiology (ABR), American Board of Medical Physics (ABMP),
Canadian College of Physicists in Medicine (CCPM)).
The sophistication of modern medical physics, as well as the complexity of the technologies
applied to diagnosis and treatment of human disease by radiation, demand this stringent approach
to becoming a member of the medical physics profession. On-the-job training simply does not
provide, with the same degree of efficiency and quality, the depth and breadth of knowledge
required of physicists entering the medical physics profession today.
Pioneers and early workers in medical physics came from traditional branches of physics such as
nuclear physics, high-energy physics, solid-state physics, etc. By chance they ended up working
in nuclear medicine, radiology or radiotherapy, and developed the necessary skills and
knowledge through on-the-job training. In addition to clinical work, they also promoted medical
physics as a science as well as a profession, and developed graduate medical physics educational
programs, first through special medical physics courses offered as electives in physics
departments, and later through independent, well-structured medical physics programs that lead
directly to graduate degrees in medical physics.
Many graduate programs are now available to an aspiring medical physicist and progression
through the four steps is feasible, albeit still somewhat difficult because of the relatively low
number of accredited academic and residency programs in medical physics. The number of these
programs is growing, however. We are now in a transition period and, within a decade,
progression through the four steps will become mandatory for physicists entering the medical
physics profession. The sooner broad-based didactic training through graduate programs in
medical physics becomes the norm, the better it will be for the medical physics profession and
for the patients the profession serves.
Rebuttal
"What does being educated for a career in medical physics mean?," asks Dr. Rogers, and answers
with two essential elements: a physicist's approach to problem solving and having research
experience. He then points out that these two elements can be obtained by progressing through a
B.Sc. degree in physics to a graduate degree in any traditional physics discipline. However, Dr.
374
Rogers ignores one additional essential element of medical physics education: having the basic
knowledge of all aspects of medical physics and a rudimentary knowledge of fields related to
medicine such as anatomy and biology. Today, this knowledge is best attained from a well-
structured academic graduate program in medical physics, rather than from on the job experience
while working as a medical physicist.
Dr. Rogers points out that AAPM awards generally go to those who have come into medical
physics from other branches of physics. This, of course, does not prove that coming into medical
physics from elsewhere is better or even equivalent to coming from a graduate program in
medical physics. It only highlights the fact that the award recipients are senior medical physicists
who entered medical physics years ago from other branches of physics. In the past, didactic
medical physics programs did not exist and medical physics and technology were far less
sophisticated than today. I predict that AAPM awards in the not-too-distant future will start
shifting to medical physicists educated in dedicated medical physics graduate programs.
Organizations offering medical physics certifications already recognize the importance of

academic as well as clinical training in medical physics by insisting on a broad basic knowledge
of medical physics during various components of the examination process. This basic knowledge
is difficult, although, as Dr. Rogers points out, not impossible, to attain on the job where one is
heavily involved with acquiring clinical experience and providing service to patients. Of course,
a physicist trained in another branch of physics can become a medical physicist; the transition is
not trivial, however, and is far less efficient than coming from an accredited graduate program in
medical physics, cross-fertilization and diversity of backgrounds notwithstanding.
Another option, not debated here yet of some relevance to the debate, is entry into medical
physics through an undergraduate B.Sc. program in medical physics. While in principle this may
give the student an advantage in a subsequent medical physics graduate program, the early
undergraduate concentration on medical physics occurs at the expense of general undergraduate
physics as well as mathematics courses. This concentration adversely affects the students'
subsequent graduate career in medical physics.
While other options remain open, presently the most efficient path to a career in medical physics
is through the well-defined and rigorous four-step progression: (i) B.Sc. in physics; (ii) graduate
degree in medical physics; (iii) residency; and (iv) certification.
AGAINST THE PROPOSITION: David Rogers, Ph.D.
Opening Statement
I have been involved with Carleton University's graduate program in medical physics since 1986.
I feel strongly that good graduate education in medical physics is valuable. I am also one of a
large number of medical physicists who joined the profession via other graduate degrees in
physics, in my case nuclear physics. When my friend Ervin Podgorsak and I were approached to
take part in this "debate," the original wording was "Medical physicists are better trained if "
rather than "educated." I would not debate the earlier proposition, which is obviously true. What
we are actually debating in this Point/Counterpoint is the difference between "educated" and
"trained."
375
What does being educated for a career in medical physics mean? It involves two essential
elements. The first is having a physicist's approach to problem solving. The second involves
having research experience to gain the ability to tackle new problems in a systematic way,
beyond solving problems in a course or textbook.
Having a physicist's approach to problem solving is a characteristic learned in any good

undergraduate physics education. A colleague who runs a large molecular biology laboratory
with a half-dozen physicists and a dozen biologists makes the point that the physicists and
biologists attack problems in completely different ways. We all recognize this fundamental
aspect of our undergraduate education as physicists.
What we learn in graduate school is how to work independently on a problem. The question is,
are we better educated if the problem is related to medical physics? There is no evidence that this
is the case. In fact, I believe that medical physics is well served by physicists from other areas of
physics joining the profession, because the variety of backgrounds is valuable. This diversity
leads to strength and robustness in the profession. Those with other physics backgrounds must be
properly trained before they work independently in a clinic, but we are debating education, not
training. Medical physics is such a broad field that even someone with a graduate degree in the
field must still learn most of the necessary specific knowledge and skills by working or training
in a clinic.
At the AAPM awards ceremony every year, most of those receiving honors and awards have
come from other branches of physics. These backgrounds can lead to very productive careers in
medical physics. This is not just a generational issue which is now different for the younger
generation. There are outstanding contributions to our field by those who were trained in other
branches of physics in the last ten years.
Note that I am not arguing that physicists from other disciplines are better educated, only that
they are equally well educated. I also recognize that a graduate degree in medical physics may be
the most efficient path for someone to enter the field, but it is not the only path. Our field is well
served by the breadth and diversity of the physics backgrounds of those entering the field. The
fact that individuals with nonmedical physics degrees will take longer to become fully qualified
to practice clinical physics should be seen as one of their personal contributions to medical
physics. This diversity of backgrounds adds a distinct strength to the field, and I would argue
strongly against restricting our discipline to those graduating from medical physics graduate
programs.
Rebuttal
I concur with much of what Professor Podgorsak has written and certainly agree that the most
efficient route for entering the profession is via a graduate program in medical physics (I will
even agree it should be accredited once Carleton's program attains that status!). But I disagree
that a well-educated physicist from any subdiscipline is limited to the field of their degree. Given
time, a well-educated physicist can pick up the specific skills required to work in most other sub-
fields, so long as individual talents are respected (i.e., most theorists are incapable of becoming
experimentalists and vice versa). Making the switch may be inefficient for the individual
involved, but all branches of physics benefit from cross-fertilization. For that reason I reject the
notion that a degree from an accredited program is an essential step to becoming a certified
medical physicist. Such a step would limit the breadth of experience of physicists entering the
profession. The examination process for certification must be adequate to assess that an
376
individual has attained the necessary knowledge in medical physics. That said, I certainly agree
that the easiest and fastest way to enter the profession is via a graduate education in medical
physics.
377
9.5. Medical physics education programs should include an in-depth

course in statistics and mathematical modeling of human anatomy
and physiology
Donald E. Herbert and Gary A. Ezzell
OVERVIEW
Medical physics is an experimental science based on data acquired with an imprecision that is
quantified by statistics. Further, the applications of medical physics are expressed increasingly
through the use of mathematical models of human anatomy, physiology, and pathology. Many
claim that understanding and using these principles is essential to professionalism in medical
physics. Others believe that these principles are peripheral to medical physics, and should not
occupy prime time in the medical physics curriculum. These differing viewpoints are discussed in
this issue of Point/Counterpoint.
Arguing for the Proposition is Donald E. Herbert, Ph.D. Dr. Herbert is Professor, Department of
Radiology, College of Medicine, University of South Alabama; Director, COM Biostatistics and
Epidemiology Core Unit, and Director, Biostatistics and Epidemiology Support Group,
Comprehensive Sickle Cell Center. He was a member of the Committee on the Biological Effects
of Ionizing Radiation of the National Academy of Sciences and a co-author of the committee's
BEIR V Report (1990). He is a Fellow of the AAPM and has chaired several AAPM committees
and task groups. He studied at Carnegie-Mellon University and Northwestern University (BS in
Physics) and at Johns Hopkins University and the University of London (Ph.D. in Physics), and
taught Physics at Colorado College.
Arguing against the Proposition is Gary A. Ezzell, Ph.D. Dr. Ezzell began practice as a clinical
physicist in 1977 in Atlanta after receiving an M.S. from the Georgia Institute of Technology. In
1979 he moved to Mount Sinai Medical Center in Cleveland. He joined Harper Hospital and
Wayne State University in Detroit in 1984, working clinically and teaching in the academic
program. He received his Ph.D. in medical physics from Wayne State in 1994. In 2000 he joined
Mayo Clinic Scottsdale as section head for physics in radiation oncology.
FOR THE PROPOSITION: Donald E. Herbert, Ph.D.
Opening Statement
Persuasive arguments in support of the proposed enrichments of medical physics academic

programs can be found among the epigrams of the 1993 AAPM Report 43: 1) "Statistics is the
theory and practice of matching theory to data." Since "matching" is what we do every day, it is
important to be good at it. Moreover, 2) "It is possible to maintain that statistics in its broadest
sense is the matrix of all experimental science and is consequently a branch of scientific method,
if not Scientific Method itself; and hence, that it transcends the application of the scientific
method in sundry fields of specialization. The scientist should know statistics as he knows logic
and formal language for communicating his ideas." [All science is either statistics or it's stamp
378
collecting! (Apologies to Lord Rutherford).] Furthermore, in our changing world medical

physicists must become 3) scientific generalists who can "practice science—not a particular
science," i.e., become more problem oriented and less discipline oriented. (The more insistent—
and more interesting —problems rarely respect disciplinary boundaries.) A knowledge of
statistics will equip medical physicists to become 4) "inference experts," i.e., "universal experts
whose specialty is not so much a subject matter as a method of inference applicable to all subject
matters." They can perform better as 5) "purveyors of the scientific method" to the medical
profession, an increasingly important role. After all, "the product of a statistical analysis is not a
posterior distribution, or a decision, or a p-value, or other summary of the data, but rather what a
statistician produces is an argument the whole argument, including the assumptions, the
logical steps involved, and the conclusions the role of diagnostics and sensitivity studies
becomes central; we need to study how the argument is affected by changes in all the
assumptions.1" Providing such arguments fall within the ambit of what can be expected of a
physicist by his colleagues in medicine and administration as we continue through A. Relman's
"Era of Assessment and Accountability2 " with Evidence-based Medicine, Technology
Assessment, Outcomes Research, the JCAHO's probabilistic definition of "quality," etc.
And surely not least are the ethical arguments: "So what is the relation between statistics and
ethics? Stated simply it is unethical to carry out bad scientific experiments. Statistical
methods are one aspect of this. However praiseworthy a study may be from other points of view,
if the statistical aspects are substandard then the research is unethical.3"
The case for including mathematical modeling (including nonlinear dynamical models)—for
insight as well as for prediction—can be made just as easily: "Philosophers and historians of
medicine identify and physicians themselves attest to two major schools of medical thought."
Realism, which "asks ontological questions, builds deterministic models and engages
in bench science," and Empiricism, which "asks epistemological questions, and offers
probabilistic models" "Probabilistic medicine eschews the questions of why, and calls up
4
the odds . " The medical physicist must be able to provide expert support and instruction in
both schools.
Our world is changing fast. "Those who do not respond to their changing world will have
decreasing influence in it.5"
Rebuttal
Dr. Ezzell presents the view that Statistics is not sufficiently important to warrant a required
course in place of the current mélange of a few ad hoc disjoint, adjunctive, albeit crucial,
techniques. But the need to apply appropriate statistical principles and methods arises as
insistently and as frequently in the routine as in the research practices of medical physics, and is
encountered as regularly by the M.S. as by the Ph.D. (although the need is too often either not
recognized—"We see what we know"—or is ignored). Moreover, it now appears that the
physicist may confront the need to teach Statistics to radiology residents for ABR examinations.
Unfortunately, Statistics is a broad and deep subject. Although fascinating, it demands

considerable study to perform at the professional level of the medical physicist. It involves not
only study design and data analysis but also the analysis of rival methods of design and analysis
and the exercise of well-informed judgment in the selection of the most defensible and robust
method, given the problem, context in which it arises, data (regrettably, data are never "given,"
379
but must be "gotten," and they are invariably biased by the presence of outlying, influential, and
missing observations) and prior information available, and purpose(s) to be served.
The most efficient way to impart such knowledge is in a formal course. A required formal course
also enforces the perceptions of both the difficulty and the importance of Statistics to medical
physics practice. A formal course also teaches what remains to be learned and thus provides an
effective beginning to the lifelong commitment to augmenting one's statistical knowledge that
distinguishes the professional scientist. Otherwise, the physicist, like the physician in Shaw's The
Doctor's Dilemma, risks drawing " disastrous conclusions from his clinical experience
because he believes, like any rustic, that the handling of evidence and statistics needs no
expertness." It happens, as a close reading of the literature attests.
AGAINST THE PROPOSITION: Gary A. Ezzell, Ph.D.
Opening Statement
There is no denying the importance of understanding the statistical principles fundamental to so

many parts of our field. Examples are ubiquitous: Poisson statistics in radiation detection, ROC
analysis in imaging, survival curves in therapy, p-values everywhere. The question is where to
put the information in the curriculum. Is it better to have a separate course, as proposed, or to
include those statistics that are relevant to each subject in the material for that unit?
The problem with the one course solution is that some of the material is so basic that it needs to
be introduced early (e.g., counting statistics in nuclear medicine), and other topics are more
naturally discussed later (survival curves). Would such a course be placed at the beginning of the
program, along with such staples as radiation interactions, displacing some other introductory
class? As a matter of curriculum design, one need only look at the web sites of CAMPEP-
approved programs to see the preferred solution: None requires such a course, although several
offer comparable electives.
This observation leads to the most persuasive argument against the proposition. Educational
programs that are designed for people who want to finish with a Master's degree and go into
clinical practice face a difficult dilemma. The number of credit hours needed for the degree
remains more or less constant, but the knowledge that must be taught keeps increasing. Consider
radiation therapy physics. Twenty years ago, students learned tissue-air-ratios, Clarkson
integrations, and wedge pair techniques. Current students still need TARs, Clarkson, and wedges,
but also need to understand pencil beam algorithms, dose-volume-histograms, and radiosurgery.
In imaging physics, entirely new modalities such as MRI and PET have developed. Teaching is a
zero-sum game in an expanding universe.
This is the key point: Requiring students to take a statistics course effectively requires them not
to take something else. For example, in the Wayne State University program, after all the
required courses are taken, a therapy-inclined M.S. student typically chooses two of three
potential electives: External beam treatment planning, brachytherapy physics, or medical
statistics. They have had some of each in the core curriculum, but now have the chance to go
deeper. Very few choose statistics, simply because the other classes are more directly relevant to
their anticipated needs.
380
The situation is very different for Ph.D. students, for whom such a course is likely to be pertinent
and necessary. A researcher must master statistical concepts and modeling tools that underpin the
techniques applied in the clinic. That such a course should be offered makes complete sense, but
to require it of all students does not.
To make such a course available to all of us, students, recent graduates and established
practitioners, makes even more sense. The proper response to expanding knowledge is life-long
learning, and so would it not be good to have a series of CAMPEP-approved, remotely-directed
classes in statistics, modeling, human-factors engineering, etc., available to the community?
Program directors take note: Build it (on the web), and we will come.
Rebuttal
Dr. Herbert argues that all science has a statistical foundation, and so perhaps a working
knowledge of basic statistical methods should be a prerequisite for any graduate work in our
field. If we require calculus, perhaps we should require statistics.
I certainly grant the importance of clear analysis in the work that all medical physicists do. Most
of us have quality assurance as a major component of our jobs, and deciding test frequencies and
action levels is inherently statistical. Most of us do measurements, and every measurement report
should have an error estimate. We certainly should be in the habit of thinking in these terms, and
we should teach our students to do so. Our publications should exhibit more statistical rigor.
Routinely we read of various factors expressed to one part in a thousand, but much less routinely
is that precision explicitly defended. Our board exams should test candidates' ability to
reasonably estimate accuracy and precision and to make justifiable decisions based on those
estimates.
Does each educational program need to have a required course in statistics, or is it sufficient to
imbue each course with the necessary statistical framework? I would argue that the latter is
crucial, and the former optional. On average, anyway.
REFERENCES
1. C. Mallows, "The zeroth problem," Am. Stat. 52, 1–9 (1998).

2. A. Relman, "Assessment and accountability: The third revolution in medical care," N.
Engl. J. Med. 319, 1220–1222 (1988).
3. D. G. Altman, "Statistics and ethics in medical research," in Statistics in Practice (British
Med. Assn., London, 1982), pp. 1–21.
4. S. J. Tanenbaum, "Say the right thing: Communication and physician accountability in
the era of medical outcomes," in Getting Doctors to Listen, edited by P. J. Boyle (Georgetown
University Press, Washington, DC, 1998), pp. 204–223.
5. R. Snee, "Discussion: Increasing the value of the statistics profession," Technometrics
40, 207–210 (1998).
381
9.6. Educational programs for imaging physicists should emphasize

the science of imaging rather than the technology of imaging
Paul M. DeLuca, Jr. and Mitchell M. Goodsitt
OVERVIEW
Some imaging physics educational programs focus on the cross-cutting principles of imaging,
with specific technologies presented as applications of these principles. Proponents of this
approach believe that it provides a solid foundation for trainees to work in any imaging field.
Other educational programs emphasize knowledge of specific imaging technologies and their
applications in the clinical setting. Advocates of this pathway feel that imaging physicists
invariably confine their practice efforts to a specific technology (e.g., x ray and CT, medical,
nuclear, ultrasound or MRI), and their educational experience should support this concentration
of effort. This controversy is the subject of this month's Point/Counterpoint.
Arguing for the Proposition is Paul M. DeLuca, Ph.D. Dr. DeLuca received a Ph.D. in nuclear
physics from the University of Notre Dame, and immediately joined the University of Wisconsin
as a Research Associate. Presently Dr. DeLuca is Professor of Medical Physics, Radiology,
Human Oncology, Physics and Engineering Physics. He served as Chair of Medical Physics from
1987 to 1998. In 1999 he assumed a role in the Medical School as Associate Dean for Research
and Graduate Studies, and his administrative role was expanded in 2001 with an appointment as
Vice Dean. His research interests have concentrated on fast neutron production and dosimetry,
determination of elemental neutron kerma factors, and application of microdosimetry to radiation
dosimetry. He currently is Vice Chairman of the International Commission on Radiation Units
and Measurements (ICRU). From 1999–2003 he served as a Chair of the Nonproliferation and
International Security (NIS) Division Review Committee (DRC) at Los Alamos National
Laboratory (LANL) and currently is a member of the LANL Threat Reduction (TR) Directorate
Program Review Committee (PRC).
Arguing against the Proposition is Mitchell Goodsitt, Ph.D. Dr. Goodsitt received his M.S. in
radiological sciences and Ph.D. in medical physics from the University of Wisconsin, Madison.
After graduating in 1982, he became an Instructor of Radiology/Assistant in Physics at Harvard
Medical School/Massachusetts General Hospital. From 1986–1992, he was an
Assistant/Associate Professor at the University of Washington. In 1992, he moved to the
University of Michigan, where he is presently Professor of Radiological Sciences. His primary
areas of research are quantitative CT, mammography, and ultrasound. He presently directs a
course on the physics of diagnostic radiology for residents and graduate students, guest lectures
in the nuclear engineering department, and co-teaches an x-ray physics/CR lab for a biomedical
engineering course. He is certified in diagnostic radiologic physics by the ABR and was recently
elected a fellow of the AAPM.
FOR THE PROPOSITION: Paul DeLuca, Ph.D.
Opening Statement
382
As a confirmed experimentalist, my first instinct is to change places with Dr. Goodsitt! In any
case, the previous 40 years of unimagined creativity in imaging science, demands an examination
of the field of medical image science. Following the 1895 discovery by Roentgen, transmission
radiography and fluoroscopy, fully conceptualized and partially developed by 1896, rapidly
reached a mature state of affairs. The next 70 years showed modest advances in image receptors,
source design, HV generators, and other aspects of image acquisition. New medical imaging
modalities developed slowly in a methodical manner, including ultrasound and radionuclide-
based imaging. By the early 1970s, however, one could sense an impending revolution.
Computer processor speeds increased at a prodigious rate, transistor gate densities increased
exponentially, and processing power put early Cray-level computing power on desktops.
Computed tomography started the onslaught of modern volume-image science. Magnetic
resonance imaging devices added enormous capability to volume imaging and complemented CT
imaging. Changes after 1980 were dramatic. High performance electronics, smart control
systems, and enormous advances in large-area, fully-digital image receptors led to a broad range
of imaging devices with ever more elegant capabilities to provide very high resolution, 4D image
acquisition with highly adaptable acquisition strategies. Finally, modalities started to fuse to
permit concurrent acquisition of physiological and anatomical information—the determination of
function.
How then shall we prepare scientists (i.e., medical physicists), to work and perform research in
this developing area? Traditionally, image science curricula were founded in the modalities, the
physics of image acquisition. They usually commenced with so-called diagnostic imaging
(transmission radiography), nuclear medicine imaging (often not including PET), ultrasonic
imaging, thence volume imaging (CT and PET), and perhaps aspects of specialized digital
imaging (e.g., DSA). While satisfactory 30 years ago, this curriculum fails to capture the
underlying common image formation concepts and mathematics. The principles of image
formation are quite general and apply to all modalities. In fact, the underlying mathematics (the
inverse problem), is widely applicable across volume imaging. This was first recognized by an
early publication of the ICRU,1 and more recently in the outstanding text by Barrett and Myers2
(2004). Casual reading of the latter's table of contents emphasizes the broad nature of the math
and statistics of image formation. With this foundation, a curriculum built on these overarching
principles can with confidence proceed to a discussion that builds on determining the underlying
biological functionality, while including the prerequisite anatomical information in the broad
context of the underlying math and physics. Modality-based discussion is presented in the context
of the interrelation amongst modalities and their concomitant ability to determine function. This
is precisely the direction of the recent recommendation of the AAPM guidance documentation.3
Rebuttal
As expected, Dr. Goodsitt and I are actually rather close in our thinking as well as our shared
concerns about learning, namely can instruction and learning realistically be bifurcated into
theory and practice without compromising understanding. It truly is a matter of degree!
However, this conundrum is more or less exactly the situation encountered in undergraduate
physics or engineering. Quite often introductory physics courses, even for physics majors, are
taught without a solid foundation in calculus, differential equations or special functions. These
courses often include electricity and magnetism or classical mechanics. In these situations, and as
correctly noted by Goodsitt, in some manner or other the underlying math is taught concurrently
with the physics! Time and time again, this process has resulted in less than adequate preparation
383
for graduate level physics—perhaps adequate for a B.S. degree, but deficient for Ph.D. level
courses. In comparison, when calculus through differential (or partial differential if possible),
special functions, and linear algebra are well understood, mechanics and electromagnetic fields
take on the beauty and symmetry that truly makes them forever understood. Coming from the
former process, I still struggle with even modestly complex electromagnetic field theory having
first learned E&M without the needed mathematics.
Even so, the contrary view, defended by Goodsitt, has clear merit when the understanding of the
imaging process is very tightly coupled to the modality under study. In fact accepting that
viewpoint leads exactly to the problem. Namely, students, after a year or so of modality-based
instruction, are now challenged to understand the broad common footings that underpin all
modalities. Frustration sets in, or even worse the student never clearly grasps the common
underlying elements of the image formation process. Image processing in astrophysics or space
science starts from the first principal approach for exactly this reason. Goodsitt makes exactly
this point when he states "When students start out in a medical physics program, many have not
yet decided which modality or modalities to specialize in this can change later in their
careers research in multimodality and multiscale imaging has a promising future. Thus, it is
beneficial for the students to learn the fundamentals of each imaging modality to a substantial
depth, because they may eventually use those modalities in their research." These remarks
embody the compelling need for a common underpinning in training and on this point we agree!
AGAINST THE PROPOSITION: Mitchell Goodsitt, Ph.D.
Opening Statement
I do not think this is an either/or proposition. Rather, I believe that to produce well-rounded
imaging physicists, the education curriculum should emphasize both the technology and the
science of medical imaging. The debate, as I interpret it, is more a choice of which to emphasize
first, the physics and technology or the generalized mathematics of medical imaging. I believe it
would be a great disservice to the majority of imaging physics students if the education programs
first emphasized the generalized mathematics and cross-cutting principles of imaging (e.g.,
impulse response functions) at the expense of the physics and technology. I base this opinion on
my experiences as a student, teacher, and researcher. There is a great diversity of skills and
backgrounds of students who enroll in medical physics educational programs (e.g., students with
undergraduate majors in physics, biophysics, bioengineering, biology, computer science,
mathematics, etc.) Having a curriculum that starts with courses on the physics and technology of
each major modality would benefit the majority of these students. First and foremost, it teaches
the students the fundamentals of each modality to a sufficient depth that the students can better
appreciate the meanings of the equations they will learn in imaging mathematics courses.
Second, in many cases the physics courses provide students with introductory and conceptual
treatments of imaging topics such as Poisson statistics, the sampling theorem, convolutions,
Fourier transforms, etc. that many of the students will need to better comprehend the far more in-
depth treatments of such topics in imaging mathematics courses. When I was a student at the
University of Wisconsin, our curriculum followed this approach, and it worked very well. Since
then, in my teaching experience, I have observed the results of the opposite ordering of courses,
wherein students first take a class devoted to generalized mathematics of imaging science. These
courses typically involve very brief introductions to topics followed by derivations of fairly
complex mathematical equations related to the topics. For example in Macovski's excellent
384
Medical Imaging Systems textbook,4 which is employed in many of these courses, 31/2 pages are
devoted to deriving the generalized transmission equation for a parallel grid:
where T( ) is the transmission at angle relative to the normal, n is an integer that takes on
values between 0 and infinity, t is the thickness, h is the height, µ is the linear attenuation
coefficient, and s is the period (= 1/frequency) of the grid strips.
All of us can appreciate the elegance of this equation and other equations that appear in this text.
The problem I have witnessed is that the students and instructors frequently concentrate on the
mathematics of imaging science to the detriment of basic principles such as knowing the purpose
of grids and their effects on image quality and patient dose. Such concepts are best taught first in
a less mathematically rigorous course devoted to the physics and technology of x-ray imaging.
When students start out in a medical physics program, many have not yet decided which modality
or modalities to specialize in. Even after they've decided on a specialty, this can change later in
their careers. Furthermore, as described at the 2003 Biomedical Imaging Research Opportunities
Workshop,5 research in multimodality and multiscale imaging has a promising future. Thus, it is
beneficial for the students to learn the fundamentals of each imaging modality to a substantial
depth, because they may eventually use those modalities in their research. Once this is
accomplished it is logical to progress to the generalized mathematics of medical imaging courses,
where as stated by Macovski in the preface to his textbook, "a formal mathematical structure is
provided, which should prove useful for the reader interested in further more detailed analysis."
Rebuttal
I hate to be the old fogey here, but what worked in medical physics education 30 years ago can
still work very well today. It just has to be updated to include new technology (e.g., DR,
multidetector helical CT, MRI, PET, image fusion, etc.) The AAPM Report 3 that Dean DeLuca
cites doesn't disagree with my thesis — it recommends for image science, "modality-driven
material as well as overall materials such as the inverse problem, signal processing, etc." The
AAPM report promotes freedom in curriculum design such as combining and redistributing
topics, but the core curriculum that is outlined is basically the same as it was 30 years ago with
the updates mentioned above. The new textbook Foundations of Image Science by Barrett and
Meyers2 that is recommended by Dean Deluca does appear to be outstanding. It contains over
1500 pages of text, with probably about as many equations, covering topics such as linear vector
spaces, eigenanalysis, singular-value decomposition, pseudoinverses and linear equations, etc. I
still fear that students using this as their first textbook in medical imaging will be overwhelmed
by the complex mathematics and lose sight of the general principles. While there may be a few
exceptional students who would do fine, the majority would be better off the old way, starting
with the basic imaging physics for each modality and ending with unified imaging theory and
mathematics.
REFERENCES
385
1. International Commission on Radiation Units and Measurements, "Medical Imaging–The

Assessment of Image Quality," ICRU Report 54, International Commission on Radiation Units
and Measurements, Bethesda, MD, 1995.
2. H.H. Barrett and K.J. Myers, Foundations of Image Science (John Wiley and Sons,
Hoboken, NJ, 2004).
3. AAPM Report 79, Academic Program Recommendations for Graduate Degrees in
Medical Physics (Revision of AAPM Report No. 44) (American Association of Physicists in
Medicine, Maryland, 2003).
4. A. Macovski, Medical Imaging Systems (Prentice-Hall, Inc., Edgewood Cliffs, NJ, 1983).
5. P. L. Carson et al., "Biomedical imaging research opportunities workshop: Report and
recommendations," Radiology 229, 328–339 (2003).
386
9.7. Engineering is preferred over physics as an undergraduate

preparation for diagnostic and nuclear medical physicists
John D. Hazle and Charles R. Wilson
OVERVIEW
Diagnostic and nuclear medical physicists are engaged in the application of science and
engineering principles to solve technical challenges in medical imaging. Graduate educational
programs for medical physicists usually cover the science (especially physics) principles, but the
applied science (i.e., engineering) principles often are not emphasized. For that reason, some
physicists prefer to hire graduates of medical physics programs who have an engineering
background at the undergraduate level. This issue is the topic of this month's Point/Counterpoint.
Arguing for the Proposition is John D. Hazle, Ph.D. Dr. Hazle is Associate Professor and Chief
of Imaging Physics at The University of Texas M. D. Anderson Cancer Center. He is a past
president of the Southwest Chapter and was recently elected as an at-large member of the AAPM
Board. Dr. Hazle actively participates in the M. D. Anderson graduate program in medical
physics and is Director of the Imaging Physics Clinical Residency Program. His research is
focused on developing temperature sensitive MR imaging techniques for monitoring minimally
invasive thermal therapies, and on developing a small animal cancer imaging research program.
Dr. Hazle has been an ABR oral examiner and was a member of the ABMP MR Physics
Examination Development Committee.
Arguing against the proposition is Charles R. Wilson, Ph.D. Dr. Wilson is Associate Professor in
the Department of Radiology at the Medical College of Wisconsin. He is head of the Medical
Physics & Imaging Science Section of the Department. He has served the AAPM on a number of
committees and currently is Chair of the Diagnostic X-ray Imaging Committee. He is certified in
Radiological Physics by the American Board of Radiology and is a fellow of both the AAPM and
the ACR.
FOR THE PROPOSITION: John D. Hazle, Ph.D.
Opening Statement
Medical imaging is nearing the end of the digital revolution. Digital solutions are available for
most imaging modalities. Ultrasound, x-ray computed tomography (CT), nuclear imaging, digital
subtraction angiography (DSA), and magnetic resonance (MR) are inherently digital. Digital
chest projection x-ray systems (using computed radiography and direct digital technologies), and
full-field digital mammography are now FDA approved. So what does this evolution have to do
with proper undergraduate preparation in medical imaging physics? In my opinion, everything.
Not so long ago, we were concentrating on using our knowledge of radiation interactions with
matter to develop detectors optimized for imaging applications. Achieving adequate image
quality at acceptable patient and personnel doses presented significant challenges. An
387
undergraduate education in physics provided the best foundation for addressing these challenges.
However, the clinical problems confronting the imaging scientist today are focused more on
optimizing system performance, and less on radiation dose. Dose is still important, but is at a
more manageable level than 10–20 years ago.
Medical imaging technologies are more mature today, and an array of digital systems is available
to replace screen-film technology. Imaging physicists are increasingly asked to address problems
created by the complex data pathways of modern imaging equipment. These problems include
the application of image filtering algorithms to achieve better signal-to-noise and/or latent image
properties (e.g., optimizing edge enhancement filters for digital chest imaging), the use of
transfer functions to convert typically linear response functions to functions matched more
closely to the human visual system (i.e., logarithmic functions for hardcopy filming or soft copy
reading on computer display monitors), and employment of post-processing algorithms to further
optimize or condense data into a presentable format (i.e., calculating functional MR maps from
thousands of source images). The proliferation of digital imaging equipment and PACS will
increase these demands. A background in electrical engineering provides a firmer foundation in
practical mathematics, signal processing and systems optimization techniques for solving these
problems. Computer science/engineering is also a significant skill requirement in support of
PACS. To be honest, computer engineering as it exists today probably does not include enough
rigorous math and physics education for clinical imaging physics.
Functional imaging is a new frontier for medical imaging. As molecular imaging agents become
more available, we will be less concerned about the shapes (morphology) of organs or tumors
and more interested in their functional properties. Nuclear imaging (PET and SPECT) and MR
will lead the way in functional imaging. A bioengineering background provides a better
foundation in physiology and biology for this imaging application.
In summary, engineering is becoming a preferred undergraduate preparation for graduate work

in medical imaging physics. Undergraduate physics programs are not preparing students as well
for graduate work in medical imaging physics, and ultimately as clinical scientists. Imaging
physics graduate students with physics undergraduate degrees seem to have more difficulty
learning physiology, systems theory, computer and communications sciences compared with
engineering students. Finally, the diminishing numbers of students entering undergraduate
physics programs is of concern. If we accept that engineering undergraduate preparations are
good for medical imaging physics, then we should forge closer relationships with undergraduate
engineering programs to promote medical imaging physics as a profession.
Rebuttal
Dr. Wilson and I agree that the practices and technologies associated with medical imaging
physics will be substantially different in five years. We also agree that undergraduate engineering
is more "focused on problem solving and practical solutions" than conventional physics
undergraduate training. However, we disagree as to whether this is an advantage or disadvantage.
I believe it is an advantage.
As we move into the next decade, the role of the clinical medical physicist will continue to
evolve as an expert technical problem solver. We are increasingly asked to dissect complex
systems and solve practical problems in digital imaging, signal processing, and systems
integration. While it is intellectually stimulating to consider the "spherical chicken," clinical
388
medical physics is very much an applied science. A significant fraction of our value to the
healthcare enterprise is our ability to develop solutions to complex problems. Our value as
researchers is not only basic technology development, but bridging the gap between the clinician
(radiologist, radiation oncologist, neurosurgeon, etc.) and the imaging equipment in support of
translational and clinical research.
In summary, physics undergraduate degrees served our generation well, and will probably serve
our successors well in the future. However, an undergraduate engineering curriculum will likely
serve future medical imaging physicists better than physics as an undergraduate degree in the
next millennium.
AGAINST THE PROPOSITION: Charles R. Wilson, Ph.D.
Opening Statement
A restatement of the proposition posed in this month's Point/Counterpoint feature is: "Given
equivalent graduate medical physics training, an individual with an undergraduate background in
engineering will be a better medical physicist than one who was a physics major." I am speaking
against the proposition because I believe that engineers and physicists approach problems
differently and that these differences are established during the undergraduate years. In my
opinion, the individual with a physics background is better equipped to deal with future advances
in medical imaging. The practices and technologies of medical imaging will be substantially
different in five years, and the individual best equipped to deal with those changes will be one
with physics training and a broader perspective of the physical world.
Graduate programs in medical physics cover the physical and mathematical principles underlying
medical imaging. Graduate courses deal with the production and detection of ionizing and
nonionizing radiation and their applications to medical imaging. Mathematical tools and concepts
such as MTF, DQE, noise, and ROC for the quantitative analysis of images are included in the
medical imaging physicist's graduate training program. Whether an individual has an
undergraduate background in engineering or physics makes little difference in his/her ability to
master this body of knowledge. In either case, students will have had courses in calculus,
statistics, physics, electricity, magnetism, mechanics, etc. However, the fundamental approach to
problem solving is very different between engineering and physics.
I have first-hand knowledge of the differences between undergraduate physics and engineering.
During my first two undergraduate years, I was an engineering student and took core engineering
courses. In my third year, I switched from engineering to physics. In engineering, the curriculum
was focused on problem solving and practical solutions. In physics, on the other hand, we
focused on understanding the physical world. We were interested in solving problems, but the
solution was not the primary goal. The contrast between the disciplines was apparent in the way
professors approached their subjects. In the engineering course on electricity and magnetism
(E&M), I analyzed many circuits and learned how to design a transformer. In the physical E&M
course, when the professor presented Maxwell equations and their applications he would often
begin by saying, "Let us postulate a spherical chicken ." While I enjoyed solving practical
engineering problems, I found the physics approach to understanding the physical world to be
more satisfying.
389
I believe the differences in approach make individuals with a physics undergraduate background
better able to find unique, innovative approaches to medical imaging challenges. When
confronted with a problem, the individual trained as an engineer may not recognize the forest for
the trees. An individual with a physics background is better able to step back and see both the
forest and the trees. It is this ability to see the larger picture that I believe makes medical
physicists with undergraduate physics training better positioned to the future challenges and
opportunities of medical imaging.
Rebuttal
Dr. Hazle argues that specific problems confronting the medical imaging physicist in image and
signal processing, PACS system administration and functional imaging require training and
background specific to the problems. He believes that electrical engineering provides a better
foundation for dealing with image processing algorithms than does an undergraduate physics
background. He also feels that biomedical engineering provides a better background than does
physics for individuals working in the field of functional imaging.
While I agree with Dr. Hazle that certain engineering backgrounds are useful for preparing an
individual to work in specific medical imaging fields, I believe that the mathematical and
computer skills of individuals with a physics background are comparable to those of persons with
an engineering background. Furthermore, I believe that the individual with a physics background
who has been trained in the scientific method has superior analytical skills and is better prepared
to handle new challenges in medical imaging. What happens when the current specific problems
in medical imaging are solved or replaced by problems with new imaging modalities? Will the
medical imaging physicist with an engineering background be versatile enough to handle the
challenges presented by new technologies? I believe that undergraduate physics training with its
emphasis on scientific methodology better prepares the medical imaging physicist to handle
current and future medical imaging challenges.
In Dr. Hazle's summary he states that imaging physics graduate students with a physics
undergraduate degree seem to have more difficulty learning physiology, system theory, etc.,
compared with students with an engineering background. In my experience with graduate
students in both medical physics and biomedical engineering the capability to grasp certain
concepts depends more upon the individual's ability rather than on the student's background and
training.
390
9.8. With the expectation that cancer detection and treatment will
occur increasingly at the molecular and gene levels, medical physics
trainees should take courses in molecular biology and genetics
Barbara Y. Croft and Colin G. Orton
Reproduced from Medical Physics, Vol. 27, No. 6, pp. 1209-1211 , June, 2000
OVERVIEW
The National Cancer Institute is making a major investment in research at the molecular and
genetic levels with the expectation that, in the future, cancer diagnosis and treatment will occur
at a more fundamental level. This occurrence is expected to replace, to a large degree, current
‗‗half-way treatment technologies‘‘ such as radiation and drugs that are extraordinarily
expensive and often less effective than desired in curing cancer. In anticipation of this
development, one could argue that medical physicists should be knowledgeable about molecular
biology and genetics through coursework taken during their training. On the other hand, medical
physics is a highly sophisticated discipline, and the curriculum for educating physicists is already
overloaded with technical courses and required training and research experiences. None of these
courses or experiences can logically be dropped to make room for new courses that anticipate
future needs but do not address present knowledge requirements for physicists. Whether
molecular biology and genetics should be added to the curriculum for medical physicists is
debated in this issue of Point/Counterpoint.
Arguing for the Proposition is Barbara Y. Croft, Ph.D. Dr. Croft is a Program Director in the
Biomedical Imaging Program of the National Cancer Institute. Prior to joining the NCI, she was
a member of the Radiology Department of the University of Virginia. She is best known for a
SPECT book and a radiopharmacy textbook. She has served on national committees concerned
with the use of ionizing radiation in medicine. Since joining the NCI, she has been working with
all aspects of funding medical imaging research, especially small animal imaging.
Arguing against the Proposition is Colin G. Orton, Ph.D. Dr. Orton is the Director of Medical
Physics and Professor of Radiation Oncology, Karmanos Cancer Institute and Wayne State
University in Detroit. He is Director of one of the first accredited medical physics graduate
programs, with over 150 M.S. and Ph.D. graduates. He has practiced radiation oncology physics
for the past 35 years. Throughout his career one of his major research interests has been
radiobiological modeling and he has taught courses in radiation biology annually to residents,
therapists, and physicists. He is a Past President of the AAPM, Past Chairman of the ACMP, and
is currently President of the International Organization for Medical Physics.
FOR THE PROPOSITION: Barbara Y. Croft, Ph.D.
Opening Statement
Medical Physics trainees should take courses in molecular biology and genetics.
391
The world is changing as molecular and cellular biologists and geneticists learn more about
processes that hold cells together, occur inside cells, and transmit properties from one cell
generation to the next. At some level all biology is chemistry and all chemistry is physics. We
are moving from an era in which biology was a descriptive science to one in which chemical and
physical processes will dominate.
The effect that this has on the medical physicist is that he/she can expect to interact daily with
biologists whose background in the physics of imaging is scant, but who wish to apply imaging
to biological models because of what it can reveal about animal models of disease. The mouse
genome is 95% similar to the human (the remaining 5% is the mouse‘s whiskers and tail and our
skin pigment, for example). Because of this similarity and because mice are small, breed rapidly,
and live their lives quickly on a human scale, mice are a convenient model for human disease,
whether it be cancer, renal or heart disease. The possibility of studying a model of human disease
through the life cycle of an individual mouse, and to use that model for testing treatments,
excites molecular and cellular biologists and geneticists. The animal can be kept intact if the
biological processes can be imaged. In turn, imaging methods used to validate mouse models and
treatment methods can be applied to human sufferers of disease. The medical physicist cannot be
a fully contributing partner with the biologist in this work unless he/she has a basic knowledge of
the principles employed.
This is not to say that the medical physicist forsakes training in his/her own field of interest in
deference to molecular and cellular biology and genetics. Instead, I suggest that too great an
adherence to the purity of physics and engineering means that the physicist will have great tools
but no co-investigators and be of little interest to others. It is very exciting to contribute to
advances in medical physics and simultaneously in medical science.
In the future, the medical physics curriculum will need to be periodically re-examined with
respect to biology as well as physics. The biological sciences will need to be integrated with the
practice of medical physics in the hospital and medical laboratory.
As for the issue of a place in an already-crowded curriculum, we have to make room for courses
that inform the trainee about the circumstances to be faced in his/her future. These are additional
opportunities for learning beyond core courses, including short courses, seminars planned around
a theme, visiting professorships, etc. Trainees seem to find time to learn the computer languages
they need, for example, even though they are rarely part of the core curriculum of a medical
physics program.
Rebuttal
Dr. Orton‘s position is that what was good enough for us when he and I were growing up is good
enough for current and future generations. It is my contention that there is a new day dawning,
not only in cancer diagnosis and treatment, but in all medical specialties that depend on imaging.
That new day is typified by a complete knowledge of the human genome, of the use of this
knowledge in the examination of normal and abnormal gene expression, and of the ability to
apply molecular biological principles in the treatment of disease.
I do see the likelihood for greatly increased use of medical imaging, and, in particular, what is
called ‗‗molecular imaging,‘‘ in diagnosis and treatment of disease. I see a great potential for the
principles of molecular biology to be used in the service of imaging for treatment planning, for
392
treatment follow-up, and in treatments themselves, in radiation therapeutic methods as well as

those based on ultrasound and optical principles.
It is said of the army that they train their recruits for the last war. We want to prepare our trainees
for diagnosis and therapeutic methods of the future. It is shortsighted for us to decide that the
current curriculum is sufficient for the challenges our trainees will meet.
And of course, I am not advocating making molecular biologists out of medical physics students,
but of educating them in the principles of molecular and cellular biology and genetics, so that
they can work with scientists who think in these terms.
AGAINST THE PROPOSITION: Colin G. Orton, Ph.D.
Opening Statement
As with any profession, knowledge is everything. Medical physics is no exception. The problem
is where do we draw the line. Life is too short to learn everything. I have no doubt that molecular
biology and genetics research will play increasing roles in cancer prevention, detection and
treatment over the next several decades. It might even be argued that some specialists of medical
physics such as radiotherapy may become obsolete if cancer can be prevented by genetic
intervention. Personally, I do not buy that argument, however, because it reminds me of over 30
years ago when I first started my career as a radiation oncology physicist when colleagues told
me I was crazy to specialize in cancer treatment now that genetic research was about to conquer
the cancer problem. Certainly molecular biology and genetics research have come a long way
since that time, but cancer incidence and mortality, and the concomitant use of radiotherapy,
have continued to increase, not decrease.
I strongly doubt that medical physicists will be less able to perform their duties if they have not
taken complete courses in molecular biology and genetics. Indeed, I believe that being required
to take such courses will be detrimental, since this will mean that other, more appropriate courses
will need to be excluded from the curriculum at a time when technological developments are
forcing medical physicists to become even more specialized than ever in order to practice
effectively. As director of a graduate medical physics program for many years, I can think of
several other courses more appropriate than molecular biology and genetics that we would like to
add to our list of required courses if they could be included in the students‘ already busy
schedules. I believe that it is our primary responsibility as educators to provide our students with
a thorough knowledge of the basic principles of medical physics and how to apply them. This
should be the major consideration in the design of courses for our teaching programs. There are
numerous peripheral topics that deserve some attention, but not complete courses.
I think molecular biology and genetics could be included in this category. In the radiobiology
course that I teach, for example, we devote one two-hour lecture to these topics. Complete
courses on these subjects are available in other departments should individual students want to
include them in their electives. However, few students have been able to take advantage of these
opportunities because they either have too little time or, more importantly, they do not have the
prerequisites in biology and biochemistry to enable them to understand, or even be allowed to
take such courses.
In summary, I believe that trainee medical physicists have enough to learn without having to take
extra courses in topics such as molecular biology and genetics that are interesting but not
393
essential to their practice, and for which they are unlikely to have sufficient background
knowledge to appreciate.
Rebuttal
I feel that I must respectfully disagree with all of Dr. Croft‘s arguments. She states that, because
medical physicists have somehow found time to learn computer languages, they will also find the
time to learn molecular biology and genetics. She is mistaken. Most medical physicists have
learned that it is smarter to leave programming to programmers. What we have done is learn to
use computers, and we have been able to do this readily because we understand the basic
mathematics behind them. Unfortunately, the same is not true for molecular biology and
genetics, because few medical physicists possess sufficient basic biology and biochemistry
knowledge to be able to benefit significantly from trying to learn these topics without extensive
preparative study in the basic biological sciences.
Dr. Croft further argues that medical physicists need to know molecular biology so that they can
collaborate with biologists to image biological processes in experimental animals like mice.
Frankly, I think most medical physicists would abandon their profession if they thought this was
their destiny.
Finally, let me respond to Dr. Croft‘s assertion that, if they do not learn molecular biology and
genetics, ‗‗...physicists will have great tools but no co-investigators and be of little interest to
others.‘‘ Dr. Croft, physicists have developed x-rays, radiology, radioactivity, nuclear medicine,
PET, RIA, CT, MRI, and IMRT, to name but a few. Some tools? Some interest?
394
9.9. The concepts of transient and secular equilibrium are

incorrectly described in most textbooks, and incorrectly taught to
most physics students and residents
William R. Hendee and Daniel R. Bednarek
OVERVIEW
Many medical physicists believe that transient equilibrium between parent and progeny
radionuclides occurs when both are decaying at the decay rate of the parent. They also believe
that secular equilibrium occurs when the half-lives of parent and progeny are greatly different,
and the activities of the two are equal. These beliefs are explored in this Point/Counterpoint from
mutually opposing points-of-view.
Arguing for the Proposition is William R. Hendee, Ph.D. Dr. Hendee received the Ph.D. degree
in physics from the University of Texas. He joined the University of Colorado, ultimately serving
as Professor and Chair of Radiology for several years. In 1985 he moved to Chicago as Vice
President of Science and Technology for the American Medical Association. In 1991 he joined
the Medical College of Wisconsin, where he serves as Senior Associate Dean and Vice President
and Dean of the Graduate School of Biomedical Sciences. His faculty appointments are Professor
and Vice Chair of Radiology, and Professor of Bioethics, Biophysics, and Radiation Oncology.
He also is Professor of Biomedical Engineering at Marquette University.
Arguing against the Proposition is Daniel R. Bednarek, Ph.D. Dr. Bednarek received his doctoral
degree from the University of Chicago in 1978 and is certified in Radiological Physics by the
ABR. He is a Professor in the Department of Radiology at the State University of New York at
Buffalo and clinical medical physicist at the Erie County Medical Center. He was a charter
member of the AAPM Commission on Accreditation of Educational Programs for Medical
Physicists and has served the ABR as an oral examiner and as a member and chair of the
diagnostic radiology physics written examination committee.
FOR THE PROPOSITION: William R. Hendee, Ph.D.
Opening Statement
In thinking about radioactive equilibrium, let's start with definitions. According to Webster's
International Dictionary: Equilibrium: a state of balance between or among opposing forces or
processes resulting in the absence of acceleration or the absence of net change; a state of
dynamic balance (as of a liquid at the boiling point) in which two or more simultaneous opposing
processes (as vaporization and condensation) proceed at the same rate and thereby cancel each
other's effects. Transient: impermanent, transitory, short-lived, momentary, ephemeral. Secular:
of or relating to a long enduring process; requiring or taking ages.
In the context of radioactivity, equilibrium defines a balance between the rate of formation of a
radioactive product, and its rate of decay. When equilibrium (state of balance) is achieved, these
395
two rates are equal, and no change occurs in net radioactivity. At equilibrium, the activity of the
product is constant, and is at its maximum level.
Transient equilibrium defines an equilibrium condition that exists for only a moment in time.
Before this moment, the product activity is growing more rapidly than it is decaying, and the net
activity of the product is increasing. After the moment of equilibrium, the product activity is
decaying more rapidly than it is growing, and the net activity is decreasing. It turns out that this
net activity decreases with an "apparent half-life" equal to the half-life of the radioactive parent,
because it is the decreasing amount of parent that causes a diminution in the activity of the
product. There is, however, nothing "in equilibrium" associated with this apparent half-life. It is
also true that at the moment of transient equilibrium, the activity of the parent and product are
equal for situations where all of the parent decays to the product. This observation in itself also
does not connote an equilibrium condition.
Secular equilibrium defines a special case of transient equilibrium in which the half-life of the
parent is very long compared with that of the product. In this case the product activity approaches
but never truly achieves equilibrium. After a time, however, the rate of change of product activity
is so small that the activity of the product appears almost constant (i.e., neither increasing nor
decreasing over time). This condition satisfies the meaning of "secular:" of or relating to a long
enduring process.
Many texts have obfuscated the meaning of equilibrium by statements such as "In equilibrium,
both activities decay with the parent's half-life." 1 "When the ratio of the activities of daughter to
parent is constant, a particular type of radioactive equilibrium exists. This is spoken of as
transient equilibrium." 2 "In the state of equilibrium the daughter radioactivity decays with an
apparent half-life of the parent radionuclide rather than its own." 3 "In transient equilibrium,
parent and daughter decay together, with the half-life of the parent." 4 "The daughter nuclide
reaches a maximum, and then achieves the transient equilibrium decaying an apparent half-
life of the parent nuclide." 5
However, some texts get it right. For example, "In the 132I–132Te example, transient equilibrium
occurs (1) at only one instant in time; (2) when 132Te reaches its maximum activity; (3) when the
activity of 132Te in the sample is neither increasing nor decreasing; and (4) when the activities of
132
I and 132Te are equal." 6 As another example, "After approximately 23 hours the Tc-99m activity
reaches a maximum, at which time the reproduction rate and the decay rate are equal and the
parent and daughter are said to be in transient equilibrium. Once transient equilibrium has been
achieved, the daughter activity decreases with an apparent half-life equal to the half-life of the
parent." 7 And some texts are ambivalent, for example, "The second segment of the (activity)
curve traces what is called the period of equilibrium, during which time the amount of daughter
nuclide decreases as the parent nuclide decays." 8
I will admit that all of this is a bit academic; most of us understand the applications of radioactive
equilibrium even if we may not define it properly. But then again, definitions in medical physics
really should be aligned closely with scientific principles.
Rebuttal
I appreciate the thoughtful and succinct position statement prepared by my colleague in response
to this month's Point/Counterpoint. There are, however, several points in his position statement
that I take issue with.
396
First, I agree that "terminology is used incorrectly when its use is contrary to a generally accepted
meaning." That is the reason for my reference to definitions for "equilibrium," "transient," and
"secular" available in Webster's International Directory. It is important that uses of words in
medical physics not wander from the established definition of those words in science and
literature.
Those of us who write textbooks know of the possibility of fallible information in such sources.
Often this fallibility is a product of extrapolating concepts and definitions across several
references. Citing several textbooks as justification for a definition is risky: the authors may
simply be reading and using each other's publications.
I take issue with the comment that the definitions of transient and secular equilibrium offered in
my position statement represent "deviant" definitions. They are "deviant" only insofar that they
attempt to return the definitions of these concepts to well-accepted meanings in the wider world
of science.
The National Council on Radiation Protection and Measurements is an invaluable resource for
establishing radiation protection guidance, and it does view its mission as "representing the
consensus of leading scientific thinking" in this arena. I am an honorary member of the Council,
having completed 16 years of service on the Council in 2002. In none of the Council discussions
in which I participated do I recall any effort to establish definitions of fundamental concepts such
as radioactive equilibrium. I don't believe the Council would extend its mission to such activities.
As stated in my position statement, the subject of this Point/Counterpoint is a bit academic. A

reader who has followed the discussion certainly will understand the concept of radioactive
equilibrium and its applications in medical physics—which is, of course, the most important
objective to achieve.
AGAINST THE PROPOSITION: Daniel R. Bednarek, Ph.D.
Opening Statement
The question of whether the concepts of transient and secular equilibrium are incorrectly
described and incorrectly taught is not one that can be answered by mathematical proof or
scientific fact. This question is answered by defining what is meant by "incorrect." I would
submit that terminology is used incorrectly when its use is contrary to a generally accepted
meaning. To use the terms transient and secular equilibrium in a manner that is inconsistent with
what is described in most textbooks would be incorrect usage. The proposition then is an inherent
contradiction since in language, common usage determines the meaning of words.
What definition could the terms transient and secular equilibrium have other than as generally
taught? I could find only two sources6,9 where these terms have been unambiguously defined
differently than in the above overview. In those references transient and secular equilibrium are
described as existing "only at a single moment in time at which the rate of formation and the rate
of decay of the daughter are exactly equal." 9 In both sources, no supporting documentation for
this deviant definition is given. In fact, this moment has been identified by Evans2 and by
Marmier and Sheldon10 as "ideal equilibrium." Both Khan11 and Andrews et al.12 have strongly
refuted the concept of single-moment transient and secular equilibrium.
397
As the proposition acknowledges, the majority of textbooks utilize the definition of transient and
secular equilibrium described in the above overview. Where does this currently accepted usage
come from? It comes from historical precedent and from acceptance by authorities in the field.
Makower and Geiger13 as early as 1912 provided a definition for transient equilibrium that is
consistent with current teaching. R. D. Evans, a recognized authority, explicitly defines transient
and secular equilibrium in this manner,2 while Andrews et al.12 give a list of ten classic
references that support "the generally accepted definition and understanding of these terms."
Furthermore, this meaning is as defined by the National Council on Radiation Protection and
Measurements,14 which has a mission to "represent the consensus of leading scientific thinking."
Even the Oxford English Dictionary15 contains this definition for transient equilibrium.
If this description of the concepts of transient and secular equilibrium is supported by authorities
and recognized experts in the field, if this meaning is broadly understood by a majority of
individuals in the discipline and is as defined in a dictionary and by a scientific body chartered by
the U.S. Congress, if it can be traced to the earliest days of the science of radioactivity, then what
reasonable justification can there be to say that it is incorrect?
Rebuttal
Dr. Hendee bases his definitions of transient equilibrium and secular equilibrium on the
definitions of the individual words in these terms as presented by Webster's dictionary. It is
entirely understandable how one could arrive at his conclusion if one only had Webster's
dictionary to use as a reference. Fortunately, the majority of textbooks and scientific literature, as
well as the Oxford English Dictionary, contain a common meaning that has historical precedent
and almost universal agreement. Although it may be unfortunate that a more literally descriptive
terminology was not originally coined to describe the process, using these terms as historically
understood is not incorrect.
In Dr. Hendee's interpretation, equilibrium exists when "no change occurs in net radioactivity"
and such equilibrium exists "only for a moment in time." Classically, transient and secular
equilibrium have referred to the state of equilibrium when the ratio of activities of progeny to
parent is constant, while that moment of no change in net radioactivity has been known as "ideal
equilibrium." Most texts do not "obfuscate the meaning" but are clear in using this classic
definition. Dr. Hendee says "some texts get it right" but only quotes two6,7 neither of which gives
supporting documentation for a definition that is contrary to most other texts. Although he quotes
some texts as "ambivalent," 8 those texts may simply reflect the confusion wrought by previous
editions of Refs. 6 and 7. It is hard to understand how a condition that exists "only for a moment
in time" or "appears almost constant" can be a better description of a "state of balance." It would
indeed be unfortunate if this proposed definition were to be adopted since we would then be left
without terms to describe what has been generally understood as transient and secular
equilibrium. These are valuable concepts that serve to communicate important relationships
between parent and progeny radionuclides.
I agree that definitions "should be closely aligned with scientific principles." However, there is
no contradiction with scientific principles in the generally accepted definitions. The principles
are the same no matter what words we choose to call them. What is important is consistency in
communication. Attempts to unilaterally redefine recognized terms simply bring about confusion
in understanding the scientific principles involved.
REFERENCES
398
1. G. Friedlander, J. W. Kennedy, and J. M. Miller, Nuclear and Radiochemistry, 2nd ed.

(Wiley, New York, 1964).
2. R. D. Evans, The Atomic Nucleus (McGraw-Hill, New York, 1955).
3. R. Chandra, Nuclear Medicine Physics: The Basics, 5th ed. (Williams & Wilkins,
Baltimore, 1998).
4. W. Huda and R. Slone, Review of Radiologic Physics, 2nd ed. (Lippincott Williams &
Wilkins, Philadelphia, 2003).
5. G. B. Saha, Physics and Radiobiology of Nuclear Medicine, 2nd ed. (Springer-Verlag,
New York, 2001).
6. W. R. Hendee and E. R. Ritenour, Medical Imaging Physics, 4th ed. (Wiley, New York,
2002).
7. J. T. Bushberg, J. A. Seibert, E. M. Leidholdt, Jr., and J. M. Boone, The Essential Physics
of Medical Imaging, 2nd ed. (Lippincott Williams & Wilkins, Philadelphia, 2002).
8. R. A. Powsner and E. R. Powsner, Essentials of Nuclear Medicine Physics (Blackwell
Science, Malden, 1998).
9. H. I. Amols and M. Bardash, "The definition of transient and secular equilibrium," Med.
Phys. 20(6), 1751–1752 (1993).
10. P. Maimier and E. Sheldon, Physics of Nuclei and Particles, Vol. I (Academic, New
York, 1969).
11. F. M. Khan, "Comments on the definition of transient and secular equilibrium," Med.
Phys. 20(6), 1755–1756 (1993).
12. H. W. Andrews, M. Ivanovic, and D. A. Weber, "Pseudoradioactive equilibrium," Med.
Phys. 20(6), 1757 (1993).
13. W. Makower and H. Geiger, Practical Measurements in Radioactivity (Longmans,
Green, and Co., London, 1912).
14. National Council on Radiation Protection and Measurements, Handbook of Radioactivity
Measurements Procedures, NCRP Report 58 (NCRP, Washington, DC 1978).
15. Oxford English Dictionary, 2nd ed. (Oxford University Press, New York, 1989).
399
9.10. Physics concepts that cannot be explained from a clinical

context should be omitted in physics courses for radiologists
G. Donald Frey and Robert L. Dixon
OVERVIEW
Radiologists should understand the physics concepts that are important to the practice of
radiology. When these concepts are presented from a clinical context, they are understandable to
most radiologists. When they are described as more abstract physics concepts, however, many
radiologists find them incomprehensible. Primarily this is because the radiologists then have to
translate abstract concepts into their clinical applications. Some believe that this translation
should be the responsibility of the physics instructor—and that topics that cannot be translated
should be deleted from the course. This issue is the topic of this month's Point/Counterpoint.
Arguing for the proposition is G. Donald Frey, Ph.D. Dr. Frey attended Canisius College in
Buffalo, NY. He received a B.S. degree in physics in 1965. He then attended the University of
South Carolina in Columbia, SC. He and Dr. Dixon were students there at the same time. He
received the Ph.D. degree in 1970. For the past 30 years he has been on the faculty of the Medical
University of South Carolina in Charleston, SC. He is presently a Professor of Radiology and
Director of Diagnostic Physics.
Arguing against the proposition is Robert L. Dixon, Ph.D., Professor of Radiology (physics) at
Wake Forest University School of Medicine in Winston-Salem, NC. Dr. Dixon is certified in
Therapeutic and Diagnostic Radiological Physics and Medical Nuclear Physics by the ABR and
has been teaching Radiological Physics to radiology residents for more than 25 years. In 1985, he
received the James L. Quinn, M.D. award for teaching excellence from his department. Dr. Dixon
currently practices and teaches in the area of Diagnostic Radiological Physics, but also has
previously practiced and taught in Radiation Oncology. He is a former AAPM president, RSNA
third vice president, and a current ABR examiner.
FOR THE PROPOSITION: G. Donald Frey, Ph.D., FAAPM, FACR
Opening Statement
There is an old story that tells of a farmer whose chickens have stopped laying eggs. The farmer
calls the agricultural school but since it is Friday afternoon only the resident physicist is available
to take the call. The farmer explains the problem. The physicist says, "I think I can help you.
First let us postulate a spherical chicken." This story reminds us that we all have preferred modes
of thinking and that the mode that is good for one discipline is not necessarily the best for
another. Experience has shown that residents have neither the time nor the inclination to learn
how to solve clinical medical physics problems directly from physics concepts, but they can
readily grasp and apply a concept when it is presented in a clinical context. The physicist, trying
to understand the pattern of scatter around an x-ray machine, would likely start with the Klein-
Nishina formula. I know from experience that this is a poor approach for residents. To make the
400
concept understandable for residents it has to be put into clinical terms with drawings of
equipment, graphics, and perhaps actual measurements. Thus I propose, in support of the
proposition, that the appropriate model for teaching resident physics is one where all material
that is taught derives from and applies to the clinical practice of radiology.
For this to take place, the physicist must make the "translation" from physics to the clinical
situation. The best teachers have the knowledge of both the physics and the clinical situation so
that they can easily make the "translation." Poor physics teaching is more often the result of a
lack of clinical knowledge, than it is an ignorance of the physics.
Teaching from a pure clinical prospective has several advantages. Having the physicist teach
from the clinical context increases the amount of material that can be covered in the limited time
available, since time is not wasted on physics principles that are interesting but not relevant. In
preparation for resident teaching, physics material that has no clinical relevance should be
removed from the course of instruction, to make room for materials that are more clinically
important and interesting. In the extreme of this position we could, in homage to Ernst Mach, ban
any concept that cannot be applied in a clinical context. This would remove much of the material
in medical physics books and examinations for residents, since the material is present because it
is interesting to the physicists who teach the material, not because it is useful or interesting to the
students.
A second advantage to the clinical approach is that the resident is more likely to retain the
material and be able to reason, at least by analogy, through equipment and image quality issues
that occur in practice.
A final advantage of this method is that it preserves the image of the physicist as possessing
valuable and occult knowledge. While resident education should prepare the residents to reason
through the common problems they encounter, they should remain aware that when more
complex situations arise they need to consult with their medical physics colleagues.
Rebuttal
I would agree with Dr. Dixon that in depth knowledge of physics is better than a superficial one,
and that the well-motivated radiology student can learn physics to a considerable depth. If the
ABR announced that radiologist candidates for the board had to be able to solve the Schroedinger
equation for a square well potential they would soon all be able to do so.
However, there are limitations on the time available for the instruction of radiology residents.
This limited time is better spent ensuring that the resident understands the relation between
physics principles and clinical applications, rather than requiring the resident to have a deep
understanding of physical principles. As an ABR examiner, I have encountered physics
candidates who have a superior in-depth knowledge of physics, but who could not apply that
knowledge to clinical situations. It is much the same with radiologists. An in-depth knowledge of
physics may broaden their cultural horizons, but if they cannot easily apply their knowledge in
clinical situations, their patients will suffer.
Dr. Dixon makes the point that a broad knowledge of physics prepares one for future
innovations. Again I agree with the principle, but I doubt that many students retain much of this
additional material. Most of my residents have taken a calculus-based physics course prior to
401
entering medical school, but after four years of concentrating on their medical education, they
retain little of that knowledge. Frequently, they even have difficulty with algebra.
Finally, we agree that the ABR ultimately controls what must be learned as a minimum. We need
to prepare residents for this examination, but we also should prepare them to meet the
requirements of clinical practice. And as medical physicists, we should encourage the ABR to
have a clinically-relevant examination in physics.
AGAINST THE PROPOSITION: Robert L. Dixon, Ph.D.
Opening Statement
In my teaching career I have been privileged to teach physics to two very bright groups of
individuals whose primary interest was not physics, viz., officers selected for the navy nuclear
power program in its early days and radiology residents. If we were to apply the premise of this
debate to the first group, we would have said that "physics concepts which cannot be explained
in terms of operation of a nuclear submarine should be omitted from the course." Fortunately,
Admiral Rickover was more enlightened than this in designing a course for these officers, and we
taught physics starting with classical mechanics, progressing through atomic and nuclear physics,
diffusion and Fermi age theory, transport theory, and finally reactor physics, with parallel
courses being taught in mathematics, thermodynamics, and electronics. I don't remember
mentioning the word "submarine" when I was describing the Bohr atom, although I could have
said there are no Bohr atoms on board a submarine (or anywhere else for that matter). My point
is that in physics one has to build a basis of fundamental ideas before one can progress to
applying physics to clinical (or engineering) problems. To try to relate every concept along the
way to a clinical situation would be a tremendous nuisance. Furthermore, a good basic
background gives students the ability to master new concepts later in their careers. For example,
in the 1970s who would have thought that nuclear spins would play any role in Radiology.
Indeed, many physicists from medical physics training programs, which stressed only the
practical, were clueless about NMR, and those with traditional physics backgrounds stepped to
the forefront.
In my experience, one can keep the attention of these bright students, even if a concept does not
appear to be clinically relevant, if it is taught in an interesting manner. One of the favorite
lectures of my students is on relativistic effects. Eyes get big when I tell them that a cocked bear
trap weighs more than the same bear trap when sprung by W/c2 where W is the work required to
cock it. Similarly, the mass of the constituent particles of a nucleus is greater than the mass of the
bound nucleus by W/c2, where W is the work required to pull the nucleus apart piece-by-piece.
Now I do advocate relating the concept being taught to a clinical situation where possible and
useful. For example, when discussing the interactions of x-ray photons with matter, I first show a
normal diagnostic radiograph, and then ask the question: "What happens if I turn off
photoelectric absorption?" I then show a similar radiograph taken with 10 MV photons
illustrating that an understanding of the photoelectric absorption process goes a long way toward
understanding the production of a good image. If a physical principle can be illustrated with an
image, then by all means it should be done since radiology residents are used to this medium. I
utilize heavily the images from the ACR teaching film file (principles of imaging module) in my
course, as well as many of my own images, e.g., contrast-detail CT phantom images made at
various dose levels.
402
A final (and practical) consideration: If the writers of the ABR exam do not also buy into the
premise as stated, then the students will not have been properly prepared. The structure of the
physics course for radiology residents is largely left to the individual physicist with only the time
allotted, the ABR exam, and student interest (attendance) as constraints or boundary conditions. I
have no doubt that many physics faculty members would welcome more guidance in structuring
such a course.
Rebuttal
Dr. Frey is evidently confused about which came first—the spherical chicken or the spherical
egg. The problem with our students is not that they can only reason in some arcane "clinical
mode" as Dr. Frey implies (they have all had pre- med courses in physical science which requires
thinking in terms of models). Instead, the problem is that they are very busy learning a lot of
things which they know they will use in their careers. Many of these students see little use in
learning any physics except to pass the board exam. It is the same with my son (the artist), who
cannot see how his high school chemistry course will have any possible application in his life
(and, I suspect that his teacher has little interest in proving such relevance). The naval aviator
must take a course in aerodynamics which contains many physics concepts and equations, when
all he really wants to do is to learn air combat maneuvering (the text book attempts to show
relevance with pictures of jet fighters interspersed in the text). It is an age-old problem for
teachers of theoretical material, not just medical physicists.
I do agree with Dr. Frey that if the medical physicist has little knowledge of the ultimate clinical
application of the subject matter, then he/she cannot be an effective teacher. I always have a "lab
session" with my students in which I take them into a fluoroscopic room and illustrate the
previously-taught concepts of contrast, noise, dose, resolution; and the dependence of these
parameters on machine settings, which the operator (radiologist) can select.
403
9.11. To prepare radiology residents properly for the future, their

physics education should be expanded in breadth and depth, and
should be more quantitative and mathematically-based
Michael Dennis and Mark Rzeszotarski
OVERVIEW
The future of radiology includes greater technical complexity of existing imaging techniques, and
extension of imaging methods into the subcellular realm of molecules, biochemical mechanisms,
and genes. To participate in this future, radiology residents should learn more about physics and
its quantitative, mathematical approach to biology. On the other hand, one can argue that
radiology must become more of a partnership between radiologists and basic scientists such as
physicists, because only through such partnerships will the discipline be able to fully exploit the
opportunities presented to it by expanding technology and basic science knowledge. In this
approach, radiologists will not need to know much physics, but they will need to know some
physicists. This controversy is the topic of this month's Point/Counterpoint.
Arguing for the Proposition is Michael J. Dennis, Ph.D. Dr. Dennis received his B.S. degree in
physics from Xavier University, followed by an M.S. in Radiological Physics from the
University of Cincinnati in 1973 and a Ph.D. in Medical Physics from the University of Texas
Health Science Center at San Antonio in 1979. He has worked in community teaching hospitals
and corporate settings and has been a diagnostic medical physicist at the Medical College of Ohio
in Toledo for the past ten years. He has served on several AAPM committees and task groups,
and has chaired the Medical Physics Education of Physicians committee for the past four years.
Arguing against the Proposition is Mark S. Rzeszotarski, Ph.D. Dr. Rzeszotarski is an Associate
Professor of Radiology and Biomedical Engineering at Case Western Reserve University. He is
certified in Diagnostic and Medical Nuclear Physics by the ABR and has been teaching medical
physics to radiology residents for twenty years, first at Mount Sinai Medical Center and currently
at MetroHealth Medical Center, both in Cleveland. He has been active in AAPM and RSNA
committees on resident education, and is the coordinator for the annual AAPM Diagnostic and
Therapy Physics Review Courses.
FOR THE PROPOSITION: Michael J. Dennis, Ph.D.
Opening Statement
The technology of radiology is expanding, and so should the technical knowledge of those
responsible for the use and interpretation of this technology. Diagnostic imaging is no longer
film-screen systems, image intensifiers, scintillation detectors and single-crystal transducers. It
now employs a wide array of sources and detectors, alternative data acquisition methods, and a
variety of techniques for image processing, display, transmission, and storage.
404
The increased technology of radiology is apparent not only by walking through the department,
but also by changes in imaging physics texts used by residents. Over the eight years between
editions of one text, for example, a 25% increase in the number of pages is found.1 These changes
reflect the growth of technologies such as PET, 4D and harmonic ultrasound, computed
radiography, digital detectors and multiple-detector spiral CT. Added to this is the expanded
knowledge required for effective use of MRI with its multiple pulse sequences, suppression and
enhancement techniques, parallel acquisition methods, and spectroscopy.
The operation of imaging equipment is becoming increasingly specialized. In this environment

the radiologist, if technologically unsophisticated, cannot control the applications of the
technology. He or she becomes a consumer of the equipment's output rather than a director of its
application and use. This can limit the clinical benefit of the technologies, and decrease the
contribution of the radiologist as an imaging specialist.
In AAPM report #64 on teaching physics to radiology residents, 2 the rationale for a physics
education program is related to: (1) the safety of patients and personnel, (2) clinical expertise, (3)
technical communication skills, (4) equipment decisions, (5) radiologists as radiation specialists,
and (6) certification and licensure. Safety issues are a growing concern, because of the high dose
levels found in CT and fluoroscopy. Clinical expertise is linked to knowledge of disease
processes, but the diagnostic task is integrally tied to the technology used to reveal these
processes. Specific technical knowledge of imaging systems facilitates the acquisition and
interpretation of images. The radiologist must be able to convincingly communicate the technical
reasons for what should be done, why, and how. The radiologist is often the primary decision-
maker in the acquisition of imaging equipment, and must be able to analyze technical capabilities
and understand what is sales "fluff." Radiologists are frequently challenged by clinical specialists
who feel that they are equally capable of image interpretation within their narrow specialty. In-
depth knowledge of the technology and its applications is one of the unique aspects in the
radiologist's training. This should be enriched rather than diluted. Also, certification and
licensure of the radiologist is one of the major assurances to the public that radiological exams
are conducted safely and knowledgably.
Regarding the physics content of the American Board of Radiology exam for radiologists, the
opinion of some examinees is that the physics exam is becoming more difficult. This seems
reasonable, because the technology of the discipline is becoming more complex. It appears that
the ABR is incorporating more practical, problem based questions in the exam in place of the rote
recall of specific facts. In preparing for this exam, AAPM report #64 recommends 100 to 200
hours of physics training. In an informal survey of nine institutions, Zink and McCollough3 found
a range from 21 to 100 hours, and an average of 66 hours, of physics taught in residency
programs, covering all diagnostic modalities. This is the equivalent of a single four semester-
hour college course, and is inadequate for the effective understanding and use of the wide range
of technologies encountered in clinical imaging.
Rebuttal
How do we teach residents with regard to the increasing complexity of diagnostic imaging?
Many programs provide considerably less than the recommended hours of instruction. The fact
that a typical resident has had only minimal math and physics makes it imperative that sufficient
time is applied to develop an understanding of diagnostic imaging technologies.
405
I agree that a solid physics, mathematics and statistics foundation is the goal, and that clinicians
should be able to recognize when and where to seek out expert assistance. They also require a
higher level of knowledge and expertise to deal with the increasing technical complexity of day-
to-day operations. The radiologist is the consultant to primary care physicians. The radiologist
must understand the application and technical specifics of the various imaging tools used, and
communicate the relative benefits and risks to his or her peers. These decisions and
communications are usually made without the availability of a physicist. This is especially true
outside of major medical institutions where diagnostic physics support is likely provided on a
part-time basis. Even in modality accreditation programs requiring physics guidance, the
radiologist has the ultimate responsibility, including technical compliance. The radiologist need
not be a physicist, but he needs a very firm technical foundation in an increasingly complex
subject.
AGAINST THE PROPOSITION: Mark S. Rzeszotarski, Ph.D.
Opening Statement
Each year, the practice of radiology becomes more complex as a result of advances in technology
and ever-increasing medical knowledge. New regulatory and accreditation requirements continue
to be imposed, placing further demands on the time available for training. Meanwhile, the
average number of radiologic examinations performed per resident continues to increase. 4 The
complexity and number of images per procedure are also increasing.
I follow the recommendations of AAPM Report No. 64, by teaching a comprehensive physics
course annually.2 Ninety hours of physics lectures are provided each year, supplemented by
lectures on continuous quality improvement, critical reading skills, using electronic media,
radiation safety, regulatory compliance, reimbursement compliance, how to conduct research,
and radiation safety lab activities. Residents also attend orientation and an annual education day
of training as well as didactic lectures on the recently revised ACGME general competencies.5,6
These lectures compete with the clinical and case conference lectures to fill the fixed number of
available time slots for didactic presentations.
My goal in teaching residents is to instill a sound conceptual basis for the physics of radiology.
While I would like to provide in-depth mathematics-based physics instruction, there is simply no
time. In addition, medical schools selectively exclude individuals who would be best suited for
advanced study in physics.7 Most medical schools require undergraduates to take a single course
in physics, and less than 20% of the schools require college level mathematics or calculus.
Undergraduates who major in physics, mathematics or engineering collectively constitute less
than 6% of all students admitted to medical school. Most medical students have poor
mathematics and physics backgrounds.
Mathematics is an essential tool for the physicist. Just as an auto mechanic relies on tools to fix a
car, mathematics provides the tools for understanding physics. While a physicist is comfortable
working with integrals and exponentials, these incite fear and anxiety in most residents.
Unfortunately, there is insufficient time to teach the fundamentals of mathematics along with the
essential physics. Instead, one is relegated to teaching physics using basic mathematics,
incorporating analogies and metaphors whenever possible to help residents understand difficult
physical principles.8 One can also use the knowledge that radiologists are visual learners. They
comprehend concepts more easily from a graph, image or sketch than from an equation. 9 My role
406
as a teacher is to explain physics concepts using methods residents can understand and relate to
in clinical practice. This takes considerable time and further reduces the breadth and depth of
coverage of all important topics.
Radiologists routinely consult with physician specialists and physicists. This partnership is
essential in today's integrated technology environment where no single individual can know
everything. In the future, radiologists will also need to include mathematicians, statisticians and
biologists in the intellectual pool of clinical consultants. The real challenge will be to provide
residents with a solid physics, mathematics and statistics foundation so they recognize the need to
seek out assistance from experts when necessary.
Rebuttal
I agree with Dr. Dennis that the technology of radiology has expanded substantially. In an ideal
world, residents should receive medical physics training which is commensurate with the
complexity of the technology. However, our clinical understanding of radiology has also
increased significantly, through both improvements in technology and a better understanding of
medicine. There is a fixed amount of time available during a diagnostic radiology residency.
Limits on duty hours, and mandated accreditation, regulatory and competency evaluation
requirements, further restrict the available time for learning. The time spent understanding
medical physics and clinical radiology must be carefully balanced to produce a well-rounded
radiologist with good clinical skills along with a solid foundation in medical physics. As
technology continues to increase in complexity and sophistication, the radiologist will have to
rely increasingly on consultant experts like medical physicists. This is the standard of care in
radiology today.
I also agree that the American Board of Radiology physics examination has become more
oriented toward clinical, problem-oriented medical physics, requiring interpretive skills beyond
rote recall of facts. This is good for radiology. We must remember, however, that the exam is a
test of minimum competency. I focus on providing my residents with a solid foundation in
clinical medical physics, rather than just enough physics to pass the boards. I teach residents how
physics principles can be used to understand and improve images. Once they comprehend these
principles, they have no problem scoring well on the boards. Teaching is performed using
qualitative nonmathematical methods that are both understandable and pragmatic. These methods
require more preparation time by the medical physicist, but the end result is improved
comprehension by the resident. They learn to apply their knowledge of physics to clinical
practice, which is the single most important outcome of medical physics training.
REFERENCES
1. J. T. Bushberg, J. A. Seibert, E. M. Leidholdt, and J. M. Boone, The Essential Physics of

Medical Imaging, 2nd ed. (Lippincott Williams and Wilkins, Philadelphia, 2002).
2. AAPM Report Number 64: A Guide to the Teaching of Clinical Radiological Physics to
Residents in Diagnostic and Therapeutic Radiology (American Association of Physicists in
Medicine, College Park, 1999).
3. F. E. Zink and C. H. McCollough, "Teaching physics to diagnostic radiology residents,"
Presented at the 83rd Annual Meeting of the Radiological Society of North America, Chicago,
IL, November 30, 1997.
4. F. A. Khan, S. Bhalla, and R. G. Jost, "Results of the 2001 survey of the American
Association of Academic Chief Residents in Radiology," Acad. Radiol. 9, 89–97 (2002).
407
5. Accreditation Council for Graduate Medical Education, http://www.acgme.org.

6. J. Collins, M. R. de Christenson, L. Gray, C. Hyde, K. K. Koeller, F. Laine, and B.
Wood, "General competencies in radiology residency training: Definitions, skills, education and
assessment," Acad. Radiol. 9, 721–726 (2002).
7. Medical School Admission Requirements, 2002–2003, United States and Canada, edited
by M. T. Moller (Association of American Colleges, Washington D.C., 2001), pp. 28–30.
8. L. V. Williams, Teaching for the Two-Sided Mind, 1st ed. (Simon & Schuster, New York,
1983), pp. 54–114.
9. J. E. Zull, The Art of Changing the Brain: Enriching Teaching by Exploring the Biology
of Learning, 1st ed. (Stylus Publishing, LLC, Herndon, Virginia, 2002).
408
CHAPTER 10
Professional Issues
10.1. Medical physicists should position themselves as institutional

resources in expanding areas such as healthcare informatics and
information networking
Jon H. Trueblood and Kenneth R. Hogstrom

Reproduced from Medical Physics, Vol. 27, No. 4, pp. 631-633, April, 2000
OVERVIEW
All healthcare institutions are faced with the expensive and technically challenging demand to
continually improve their informatics capability and their infrastructure for information
networking. The demand includes the need to provide Electronic Medical Records and Picture
Archiving and Communications Systems (PACS), and their integration with Radiology (RIS)
and Hospital (HIS) Information Systems. In many institutions, one of the problems in meeting
this demand is the lack of skilled personnel who understand both the technical and the clinical
aspects of information systems and networks. Medical physicists have this understanding, and
could be a major resource in helping institutions meet their informatics and networking needs.
But assuming this responsibility would divert medical physicists from their more traditional roles
in diagnostic and therapeutic radiology, where they also have unique responsibilities. Whether to
offer to help resolve information systems and networking challenges presents a dilemma for
medical physicists. How this dilemma might be resolved is examined in this issue of
Point/Counterpoint.
Arguing for the Proposition is Jon Trueblood, Ph.D. Dr. Trueblood is Professor and Director of
the Medical Physics Section of the Department of Radiology at the Medical College of Georgia.
Dr. Trueblood is the past Chair of the AAPM Education Council and is currently Vice Chairman
of CAMPEP and Chairman of the ACR/CMP Committee on Physics Education. In 1993 he was
co-director of the AAPM School on the topic of ‗‗Digital Imaging.‘‘ For more than a decade he
was co-director of the AAPM annual Diagnostic and Therapy Physics Review Courses. He has
been instrumental in establishing the Medical Physics Continuing Education Credit (MPCEC)
Program under the auspices of CAMPEP. In 1998 Dr. Trueblood received the AAPM Award for
Achievement in Medical Physics.
409
Arguing against the Proposition is Kenneth R. Hogstrom, Ph.D. Dr. Hogstrom is Professor and
Chairman of the Department of Radiation Physics in the Division of Radiation Oncology at The
University of Texas M. D. Anderson Cancer Center, where he holds the P. H. and Fay Etta
Robinson Professorship in Cancer Research. Dr. Hogstrom has over 25 years experience in the
research and development of dose algorithms, PACS, and other software tools used in radiation
therapy. He is director of a medical physics graduate education program, member and previous
chair of the CAMPEP residency education program review committee, and AAPM President for
2000.
FOR THE PROPOSITION: Jon Trueblood, Ph.D.
Opening Statement
Outside of reviewing radiation therapy treatment charts, medical physicists, in general, have not
been significantly involved in the business of patient medical records (charts). However, along
with the transition to filmless radiology, the analog methods used by medical record departments
are also rapidly making the transition to electronic (digital) records. The process of integrating
the patient‘s electronic medical records and digital medical images has not only generated a huge
hospital information software industry but is also resulting in the formation of institutional wide
healthcare informatics and information networking service departments.
The expertise included in hospital informatics departments is necessarily multidisciplinary in

nature and generally includes some combination of computer system, database and network
engineers, telephone/cable technical personnel, business and technical managers (coding, typing,
billing, etc.), representative specialty physicians, and, in some academic institutions, a medical
informatics scientist. Frequently, the medical physicist is not included in the initial constitution
of such medical informatics service groups. This is unfortunate because, within the
aforementioned typically included specialists, who among these specialists has a perspective on
the content, format, communication and display requirements, and storage and retrieval
requirements of the following medical record entities: medical images obtained from the digital
imaging modalities (MR, US, CR, CT, etc.) and from the digitization of film and analog video
displays (fluoroscopy, video cameras, etc.); 3-D visualization images created to assist radiation
therapy, surgery and diagnosis; and CT simulation and electronic portal images? Who among
these specialists understands image quality issues in the context of lossy compression and
greyscale and color display monitor characteristics? Which specialists understand how these
images are acquired and how they need to be distributed? Which specialists are most aware of
the DICOM standards? These rhetorical questions represent an abridged list of digital medical
image entities which are outside the expertise of most medical informatics groups.
It should be obvious that, at a minimum, a medical physicist should be included as a consultant

to the institutional group charged with oversight of medical informatics. The extent to which
such duties detract from the more traditional role of the medical physicist will be situationally
variable. Even if it requires adding medical physics personnel, is it a bad thing for the institution
to need this service from the medical physicist? What other health professional possesses the
knowledge to properly integrate digital medical images into the patient medical record?
Rebuttal
My opponent has provided an excellent referenced definition of academic medical informatics

and information networking. Note that the proposition states that the medical physicist should be
410
an institutional resource in these expanding areas of healthcare. This does not mean that the
medical physicist need be an expert in the academic discipline of medical informatics or
computer networking. In the context of my opponent‘s radiotherapy accelerator analogy, medical
physicists have certainly been a resource in the design requirements for this sophisticated
instrument without being expert in the fields of electronic circuit design, mechanical and
electrical engineering or computer device control technology. Just as the medical physicist has
been the best resource for advice on the performance and functionality requirements in the
development of the accelerator, the medical physicist is the best institutional resource for
consultation on the performance and functionality of medical image acquisition, display, archival
and network systems. The important question is ‗‗Do the medical informatics professor and the
computer network engineer possess more knowledge about radiology and radiotherapy medical
information and images than the medical physicist?‘‘ If the answer is no, then
the medical physicist who does not become an institutional resource in these expanding areas of
healthcare has missed an important professional opportunity.
AGAINST THE PROPOSITION: Kenneth R. Hogstrom, Ph.D.
Opening Statement
My position is that the scope of healthcare informatics is enormous and an area in which the
medical physicist has little to no training, and hence, one in which medical physicists should not
assume responsibility. Fueled by advances in computer technology, medical informatics is the
logic of healthcare, and according to Coiera,1 the role of medical informatics is to help develop
solutions as to (1) how rational structures can be used to specify how clinical evidence is
applied to patient care, (2) how organizational processes and structures can be used to minimize
use of resources and maximize patient benefit, and (3) how tools and methods can be developed
to best achieve these roles. Information networking, which in the present context can be thought
of as a collection of computers connected together and sharing data and programs, is a tool of
healthcare informatics.
Undoubtedly, through research and development the medical physicist has contributed to the
technical sophistication and functionality of picture archiving and communication systems
(PACS) in diagnostic imaging2 and radiation oncology.3 Nonetheless, the medical physicist
should not be responsible for the development, operation, and maintenance of a PACS, which
requires a well-staffed group of computer informatics professionals that have adequate goals and
resources. Furthermore, should medical physicists perform such duties related to PACS and
information systems, then they should have a joint title indicating their increased responsibilities,
and that portion of their effort should be recognized as nonmedical physics for staffing
determinations.
The medical physicist is responsible for acquisition, commissioning, procedure planning, quality
assurance, and safety of radiological equipment. The breadth and complexity of this has
increased significantly in recent time due to the influx of new technology for patient care, e.g.,
MRI, SPECT, PET, digital imaging, and functional imaging in diagnosis and IMRT, gated
therapy, image guided therapy, and stereotactic irradiation in therapy. Therefore, the medical
physicist should participate in acquisition, commissioning, and quality assurance of image
quality, spatial linearity, and functional operations of the PACS, i.e., those functions for which
the medical physicist is typically responsible, but not in its development, operation, and
maintenance. A similar analogy in therapy has been the radiotherapy accelerator. Radiotherapy
accelerators are highly sophisticated medical instruments
411
whose developments were stimulated by physics and whose uses have been improved through
medical physics research and development. The medical physicist assists in acquiring,
commissioning and providing ongoing quality assurance of the dosimetric properties of the
machine. However, radiotherapy accelerators are developed by engineers, manufactured by
industry, and operated and maintained by other non-medical physicist professionals.
Rebuttal
The questions raised by Dr. Trueblood are typical of most emerging medical technology, and I
concur that it would be acceptable for a medical physicist to be a consultant to an institutional
group responsible for medical informatics. Both of us clearly recognize the contributions and the
knowledge of the medical physicist in digital technology applied to radiology. However, in
response to Dr. Trueblood‘s question, it could be a bad thing for the medical physicist to become
overly involved and responsible for medical informatics. Historically, too often the medical
physicist has had to solve ‗‗computer problems,‘‘ that were not related to medical physics.
Although such actions may benefit the institution, and even provide short term recognition of
medical physicists, they are seldom accompanied by increased medical physics personnel, and as
a result, distract from other duties of the medical physicist. Therefore, I will assert my earlier
point that if medical physicists perform such duties, then their effort should be recognized as
non-medical physics for staffing determinations.
Medical physics is a maturing profession whose demands are increasing because of increased
patient demand, more sophisticated technology, and an increased range of applications to
medicine. Because of this, physicians are becoming more dependent upon medical physicists.
Therefore, we need to focus on refining our proficiency as practicing professionals and not be
expanding into areas that are clearly not medical physics.
REFERENCES
1. E. Coiera, Guide to Medical Informatics, the Internet, and Telemedicine (Chapman &
Hall, London, 1997).
2. H. K. Huang, PACS, Picture Archiving and Communication Systems in Biomedical
Imaging (VCH Publishers, Inc., New York, 1996).
3. G. Starkschall, S. Bujnowski, X-J. Li, N. Wong, A. Garden, and K. Hogstrom, ‗‗A
picture archival and communications system for radiotherapy,‘‘ Proceedings of the International
Symposium on 3-D Radiation Treatment Planning and Conformal Therapy (Medical Physics
Publishing, Wisconsin, 1993), pp. 411–420.
412
10.2. Medical physicists should assume PACS responsibilities
James R. Halama and Walter Huda

OVERVIEW
The implementation of Picture Archiving and Communications Systems (PACS) is a challenging

and costly project being addressed in some form by every radiology department. Some physicists
believe that this implementation presents an obligation and an opportunity for medical physicists
to increase their contribution to the clinical service of the department. Other physicists feel that
PACS is outside the knowledge and skill set of medical physicists, and would distract them from
their responsibilities in quality control and equipment monitoring. These opposing points of view
are debated in this Point/Counterpoint.
Arguing for the Proposition is James R. Halama, Ph.D. Dr. Halama is an Assistant Professor of
Radiology/Section of Nuclear Medicine at the Stritch School of Medicine at the Loyola
University Medical Center. Dr. Halama is currently the Chair of the Nuclear Committee of the
AAPM Science Council. He has spent many years working with image processing techniques in
nuclear medicine, and pioneered bulletin board communication methods on the world-wide-web
through development of the Loyola University Nuclear Information System (LUNIS). He
currently manages the radiology imaging network, and serves as chairman of the PACS selection
committee at Loyola.
Arguing against the Proposition is Walter Huda, Ph.D. Dr. Huda obtained his Ph.D. in medical
physics at the Royal Postgraduate Medical School (Hammersmith Hospital) at the University of
London. Dr. Huda worked for five years at Amersham International, a commercial company
specializing in radioactive products. In 1982 he moved to North America, and worked in medical
physics at the Manitoba Cancer Treatment and Research Foundation (Winnipeg) and the
University of Florida (Gainesville). Dr. Huda is currently a Professor of Radiology at SUNY
Upstate Medical University (Syracuse), and Director of Radiological Physics. His research
interests are in medical imaging and radiation dosimetry, and he is certified in Diagnostic Physics
by the CCPM and ABMP.
FOR THE PROPOSITION: James R. Halama, Ph.D.
Opening Statement
In many radiology departments today a gap exists in integrating information technology and
PACS. In a very short time, all radiological imaging will be digital and connected within a
PACS. Departmental administrators understand cost and workflow issues; information
technology personnel understand computer networks, databases, and message exchange;
technologists know how to produce images, but lack the technical understanding of image
production; and few physicians look at image quality beyond what is presented on the video
screen. Because of their technical training, orientation, and understanding of images from
production to presentation, medical physicists are the persons who can bridge all of these gaps.
413
Medical physicists have the opportunity and the responsibility to adapt to the new PACS
environment.
The starting point in gap bridging is having an understanding of images and the data that are
stored along with images. It is clear, as previously argued by Trueblood and Hogstrom, 1 that
medical physicists are the specialists who have a perspective on image content, format,
communication, display, storage and retrieval. Medical physicists understand image quality in
the context of glossy image compression, and grayscale and color display characteristics.
The DICOM standard, that enables PACS to exist, is open to interpretation by both imaging and
PACS vendors. Critical nonimage data elements in the DICOM headers may not be interpreted
correctly across vendor platforms, and may not be stored at all in a PACS. Increasingly, PACS
are being used as the data repository for imaging modalities. Expertise is needed not only to
solve cross-platform connectivity issues, but also to validate the vendor's interpretation of
DICOM data elements. This expertise is critical because of the increasing demand for 3-D
visualization tools in PACS workstations, the emergence of image-guided therapy, and image
fusion between CT, MR, and PET. DICOM, although viewed by many as complete, is still an
emerging standard, and validation of procedures will be required for many years to come.
In planning for, or extending the reach of PACS, questions arise almost daily concerning image
storage and bandwidth requirements for transferring images to and from on- and off-site
locations. The input of a medical physicist in departmental planning and PACS committees will
help the department select the proper PACS and network configurations. In addition, good
technical input often keeps the department from over-specifying performance criteria.
Assumption of PACS responsibilities by physicists enhances the quality of patient care and helps
the radiology department operate more efficiently. PACS responsibilities, however, require that
physicists are knowledgeable in this area. We must provide more continuing education programs
for physicists who are being called upon to work with PACS.
Rebuttal
Unfortunately, PACS is being narrowly viewed as merely a tool for improving the operational
efficiency of Radiology departments. We must remember that PACS archives, transmits, and
presents images to physicians. In Radiation Oncology, PACS is used to archive, transmit, and
present radiation treatment plans. Medical physicists should, at a minimum, have sufficient
knowledge of DICOM, computer networks, interfaces to modalities, and databases to ensure that
the integrity of image and related data are maintained from production to presentation. This
responsibility for image and data integrity is being coded into federal regulations. For example,
the revised NRC Part 35.41 states "For any administration requiring a written directive, the
licensee shall develop, implement, and maintain written procedures to provide high confidence
that at a minimum address verifying that any computer-generated dose calculations
are correctly transferred into the consoles of therapeutic medical units authorized by § 35.600.2
DICOM RT was developed recently to standardize the transfer of radiation treatment plans to
therapeutic devices. I would be very uncomfortable in suggesting that a computer scientist take
on this responsibility.
Yes, there will be individual medical physicists who have sufficient knowledge to become PACS
administrators and oversee the total PACS operation. The fact remains that most medical
physicists, at some time in their careers, will be called upon to perform PACS responsibilities
414
insofar as the integrity of image and related data is concerned. The AAPM should initiate
continuing education programs to raise the level of competence of medical physicists in PACS,
and medical physics training programs should incorporate PACS into their curricula. By offering
these programs, medical physicists will not need to seek education and training from other
societies and professional groups.
The underlying issue is: Do medical physicists remain the imaging specialists? Or do they hand
over increasing bits and pieces of new technology to computer scientists who have no knowledge
of the specialized requirements of Radiology?
AGAINST THE PROPOSITION: Walter Huda, Ph.D.
Opening Statement
There will be individual medical physicists who may develop an interest in PACS and acquire
the necessary skills to assume PACS responsibilities. My contention, however, is that most
medical physicists working in radiology departments will not and should not assume PACS
responsibilities. Diagnostic physicists currently have their hands full dealing with issues of
radiation dose and image quality In most radiology departments, responsibility for PACS should
be assumed by computer scientists with informatics expertise.
Medical physicists usually have undergraduate degrees in physics, and they gain their medical
physics training with an M.S. and/or a Ph.D. level degree. A typical training program in
diagnostic physics includes interactions of radiation with matter, radiation biology and
protection, and medical imaging for the major modalities currently encountered in radiology (i.e.,
radiography, fluoroscopy, DR & CR, CT, Ultrasound, MR and Nuclear Medicine). Although
virtually all physicists are knowledgeable about computers, most have little direct experience of
the DICOM standard, interfaces to imaging modalities, network issues, or databases.
Certification boards such as the American Board of Radiology and the Canadian College of
Physicists in Medicine do not cover PACS in a substantial way in their written and oral
examinations.
The advent of PACS is dramatically changing the way that Radiology departments operate. With
PACS, radiographic images are digitally acquired, processed, transmitted and archived. In the
near future, the radiologist's report will be dictated using voice recognition dictation systems.
Radiology Information Systems (and Hospital Information Systems) will then permit the end
product of the radiological examination (i.e., report + images) to be available on line to referring
physicians in a matter of minutes or hours. This is a huge improvement over film-based imaging
systems where the median time to produce a verified report is measured in days! PACS is
primarily about improving the operational efficiency of radiology departments, which is not the
domain of diagnostic medical physics. To introduce PACS into radiology requires detailed
technical knowledge about DICOM, interfaces, networks, and databases.
Imaging physicists serve an important role in monitoring technological trends such as the recent
introduction of multislice CT scanners. Keeping the department at a state of the art level requires
joining appropriate societies (e.g., American Association of Physicists in Medicine), attending
relevant meetings (e.g., SPIE Medical Imaging), and reading pertinent journals (e.g., Radiology).
To be knowledgeable about PACS requires joining the American Medical Informatics
Association, attending the Healthcare Information and Management Systems Society (HIMSS)
415
meeting, and reading the Journal of Healthcare Information Management. Most medical
physicists are not even aware of the existence of these entities, and are certainly not active
participants. Therefore they are not equipped to assume the primary responsibility for PACS in a
radiology department.
Rebuttal
My colleague makes the claim that a gap exists in integrating information technology and PACS;
expert personnel are required to address cross-platform connectivity, validate the vendors
interpretation of DICOM data elements, and tackle issues of communication, storage and
retrieval requirements. All of these PACS issues are important. But they are beyond the expertise
of most diagnostic physicists. The AAPM is the professional home of medical physicists,
Medical Physics is the journal we read, and there is an annual meeting that we normally attend
each summer. A review of the contents of the journal and annual meeting shows that there is
virtually no mention of PACS topics. It is doubtful that more than a small percentage of
diagnostic medical physicists would feel comfortable in bridging the gap between information
technology and PACS.
I fully agree that medical physicists have a special understanding of dose and quality issues in
medical imaging. There is little doubt that we will continue to have an important role in
providing 3-D visualization tools, image guided therapy and image fusion. However, as I argued
in my opening statement, PACS is primarily about the improvement of the operational efficiency
of radiology departments. Efficiency will be achieved by the effective management of images
and reports, and integration of PACS with RIS and HIS systems. Individuals who do venture into
the PACS arena will most likely have little time left to devote to the traditional diagnostic physics
tasks (specification, acceptance testing, quality control, and clinical use). The success of a
radiology department depends on a team effort consisting of radiologists, physicists,
technologists, manufacturers, as well as PACS personnel.
It is of interest to contemplate to what extent medical physicists who operate outside of the larger
academic medical centers will become involved with PACS. A substantial percentage of
radiology centers do not employ full time physicists, and many diagnostic physicists operate in
private practice to provide consultation services to the medical imaging community. To guide a
radiology group through the difficult transition from a film based enterprise to the all digital
world requires an individual knowledgeable about PACS issues, and not a medical physicist per
se. I believe that most private medical physics groups get very little PACS business, and I doubt
that they are positioned to successfully move into this area, unless they expand by recruiting
computer scientists and information technologists with the necessary expertise.
REFERENCES
1. J. Trueblood and K. R. Hogstrom, "Medical physicists should position themselves as

institutional resources in expanding areas such as healthcare informatics and information
networking," Med. Phys. 27, 631–633 (2000).
2. United States Nuclear Regulatory Commission, 10 CFR Parts 20, 32, and 35 RIN 3150-
AF74 Medical Use of Byproduct Material, Subpart B—General Administrative Requirements,
Section 35.41 Procedures for administrations requiring a written directive, April 24, 2002.
416
10.3. It is important that medical physicists be involved in the

development and implementation of integrated hospital information
systems
George C. Nikiforidis and George C. Kagadis

OVERVIEW
It is becoming increasingly difficult to operate a hospital without an integrated hospital

information system (HIS) that streamlines the collection, management, and distribution of
information necessary for health professionals to function efficiently. Such electronic data
processing systems are used to manage information from such diverse sources as laboratories and
radiology departments. Laboratories typically do not have staff that is highly knowledgeable
about data collection and exchange. On the other hand, radiology departments often have medical
physicists experienced in managing PACS systems, and it has frequently been these medical
physicists who have evolved into local HIS experts. Some hospitals realized the importance of
hospital information systems early on and employed dedicated medical informatics specialists for
the development and implementation of their information systems. In these hospitals, medical
physicists have often been only marginally involved in information technology, and some have
suggested that this marginal involvement is detrimental to our profession. They have
recommended that medical physicists should spend more time and effort developing their
expertise in medical informatics so as not to relinquish this responsibility. Others have suggested,
however, that medical physicists do not, and should not, take on these extra duties because they
have neither the training nor the time for such endeavors. This controversy is the topic of this
month's Point/Counterpoint debate.
Arguing for the Proposition is George C. Nikiforidis, Ph.D. Dr. Nikiforidis received his Laurea in
Physics and his M.Sc. in Atomic and Nuclear Physics both from the University of Milan, Italy in
1973 and 1980, respectively. He received the Ph.D. in Medical Physics from the University of
Patras, Greece in 1981. He is currently Professor of Medical Physics and Director of the
Department of Medical Physics at the University of Patras, where he is also the Dean of the
School of Health Sciences and director of the postgraduate course on Medical Physics. He has
been the principal investigator or been involved in a variety of national or European research and
development projects.
Arguing against the Proposition is George C. Kagadis, Ph.D. Dr. Kagadis received his Diploma
in Physics from the University of Athens, Greece in 1996 and both his M.Sc. and Ph.D. in
Medical Physics from the University of Patras, Greece in 1998 and 2002, respectively. He is
currently a Research Fellow in Medical Physics and Medical Informatics at the Department of
Medical Physics, University of Patras. He is a Greek State Scholarship Foundation grantee, and a
Full AAPM member. He has been involved in European and national projects, including e-health.
His current research interests focus on medical image processing and analysis as well as studies
in computational fluid dynamics. Currently he is a member of the AAPM Imaging Informatics
Subcommittee and an Associate Editor of Medical Physics.
417
FOR THE PROPOSITION: George C. Nikiforidis, Ph.D.
Opening Statement
Hospital information systems have become key elements of the infrastructures of modern
hospitals in the developed world. Their role in a hospital's main function of providing health
services is twofold: they act both as a tool for managing the entire spectrum of activities within
the hospital organization, and as mechanism for integration of newly acquired data and
knowledge into clinical routines.1,2,3,4,5 With regard to the former, a medical physicist (MP) has
little to contribute. Information technology scientists are natural protagonists in fields where
technological development in hardware and software is so rapid. The latter, however, forms a
new dimension of development for an HIS. Its dynamic character lies in the process of
continuously including new types of data emerging from new diagnostic methodologies. This
continuous updating of HIS constitutes a challenge for modern medicine. First, technological
advances offer the opportunity to perform diagnostic examinations at increasingly subtle levels
of pathology; from the organism to the tissue, the cell and—nowadays—the molecule.
Extrapolating present trends, it is quite probable that in the ensuing years a diagnostic procedure
might simultaneously include data originating, for example, from multislice CT, PET,
microarrays, and proteomics.6,7 Second, a new need is formed for combining data optimally and
maintaining a holistic approach to clinical problems. Approaching the clinical problem from
diverse diagnostic aspects should strengthen our perception of the problem at hand without
leaning towards technological fetishism.
Medical physicists can play significant roles in this framework. Their firm scientific knowledge
of the physical principles of data acquisition from a number of diverse diagnostic modalities
(making use of ionizing as well as nonionizing radiation) can assist them not only in filtering out
sources of noise convolved to the measurements, but also in bringing out the additional value of
specific diagnostic methodologies over others. 8 As the hospital environment becomes more
molecular, new entities such as microarrays and molecular imaging are emerging. The
fragmentary character of information calls for integrative initiatives that will combine the
scientific backgrounds of basic sciences such as physics, chemistry, and biology.
Exploiting their background knowledge of the mechanisms of data acquisition, medical

physicists are able to perform data reduction in the form of extracting characteristic features from
the data, i.e., variables that contain summaries or inferential information from highly complex
raw data. The MP can thus be actively employed in the process of learning from data originating
from heterogeneous sources of information and aiming at various groups of end users. Medical
decisions are intrinsically of statistical character, and pursuing plausible hypotheses involves
inference from complex data structures.
HIS can become an indispensable tool for such tasks, and MPs can act as moderators for diverse
medical needs met through inference inside the hospital. Their role among end users can provide
better exploitation of the knowledge gathered in a medical setting, and can serve as an effective
inter-science collaboration between physicians and physicists. I strongly believe that by playing
an active role in the ever-evolving HIS environment, MPs can exert a positive influence in
hospital organization. This gives MPs the opportunity to offer better services to the hospital and
to patients.
AGAINST THE PROPOSITION: George C. Kagadis, Ph.D.

418
Opening Statement
The development and implementation of integrated hospital information systems is a very

interesting issue. Getting involved with this task requires a strong background in medical
informatics science and is highly work-intensive. My position is that medical physicists have
neither an adequate background in this science nor the time to devote to it, and hence should not
assume such a responsibility. Medical physicists usually have an undergraduate education in
Physics and an advanced degree in medical physics. According to AAPM Report Number 79,
there should be a core curriculum in which all medical physics trainees should be well grounded.
In addition, there should be subsequent training in the more specific aspects associated with
subspecialties such as imaging or radiation therapy.9 Medical physicists have a variety of
obligations such as acceptance tests, shielding design, radiation dosimetry, treatment planning,
image quality, etc. Being either a diagnostic or therapy physicist demands devotion and
responsibility in everyday practice. Medical physicists' obligations have significantly increased
recently with the advent of new technological advances such as multidetector CT, PET, IMRT,
MLC, and Gamma Knife, as well as hybrid machines CT-SPECT, MR-PET, etc.10 Thus
specialization and dedication are crucial if medical errors are to be avoided.
Development and implementation of an integrated HIS alter the way clinical enterprises operate,
and require specialists who have a solid background in medical informatics. These individuals
should have in-depth training in PACS, database design and management, interfaces to imaging
modalities, communication protocols DICOM, TCP/IP, as well as higher-layer HL7 protocols,
etc. An HIS will make the outcome of a radiological, laboratory, or other physical examination
readily available to referring physicians. A well-developed and implemented HIS will increase
the operability and consequently the performance of the healthcare enterprise. 11,12,13
The current medical physics curriculum does not provide sufficient tools to enable medical
physicists to have a key role in the development and implementation of an integrated HIS.
Additionally, medical physicists have little (or at best, fragmentary) knowledge of medical
informatics issues, and thus any attempt to be involved in the development and implementation
of an integrated HIS would distract them from their specialized work and thus decrease
productivity and efficiency in their daily clinical practice. The partnership of medical physicists
with radiologists, other physicians, and medical informatics specialists is essential in an
integrated HIS, where each individual has a discrete role and cannot have knowledge of
everything.14
Rebuttal: George C. Nikiforidis, Ph.D.
It would be unrealistic not to admit the need for greater specialization of MPs in individual
aspects of medical physics so that they can embrace the new technologies and methodologies that
enter our everyday practice in the hospital. If, however, this were not accompanied by efforts to
link the physicist's routine work to the broader goals of modern medicine, then the very act of
specializing would cause their marginalization. Their role would become ancillary and MPs
would undoubtedly narrow their scientific role to one of simply striving to follow continuous
technological advancements.
Hospital information systems are the means for a solid attachment of MPs to the roadmap of
advancement in medical science, as they are the key factor for the integration of scientific
knowledge. Giving MPs the opportunity to be involved in the development and implementation
of an HIS makes them active contributors to this roadmap. Registration and fusion of medical
419
images constitute good examples for such integration. Taking advantage of the HIS
infrastructure, the information is combined to produce fused images. Interpreting them comprises
a challenging task for the clinicians, since they are required to evaluate a new, unfamiliar
representation of the information. It calls for collaboration among experts, opinion exchange and,
frequently, argumentative reasoning. The MP, aware of the physical principles involved in the
image acquisition mechanisms, can effectively assist the clinicians in this task.
Being able to take part in such schemes undoubtedly requires proper education in fields that lie at
the interface between scientific disciplines. Knowledge of the mechanisms of novel acquisition
methodologies, as well as aspects of computer science, is only a part of the technologies
necessary to allow MPs to perform information integration and implement statistical learning
procedures. Armed with such qualifications, the role of MPs as knowledge facilitators would be
enhanced. This would promote the importance of medical physicists in the hospital environment.
Rebuttal: George C. Kagadis, Ph.D.
I agree with my colleague that hospital information systems have become key infrastructures of
modern health care enterprises. I also agree that the role of the current medical physicist is
expanding to new fields according to medicine's development. However, I disagree with his
statements that the medical physicist can act as a moderator for the diverse needs for medical
inference in the hospital setting. Such an approach would add more tasks to the daily workload of
medical physicists and would divert them from their important duties. They have to be efficient
in their daily clinical work as well as stay up-to-date with new technological developments in
either imaging or radiation therapy. These tasks demand devotion in order not to decrease
productivity and efficiency.
I also agree that active participation of medical physicists in either the development and/or
implementation of an integrated HIS may positively influence both their status (strategic role)
and the clinical enterprise function. On the other hand, medical physicists are evaluated and
acknowledged for their services as medical physicists, and any possible recognition with regard
to HIS services might just be transitory. Additionally it is not likely that this new involvement
would lead to an increase in medical physics department personnel. This means that if medical
physicists do assume HIS responsibilities in either development and/or implementation, they will
inevitably have less time to dedicate to their traditional tasks and thus compromise their quality
of service or, in the worst case, leave them more vulnerable to making errors.
For these reasons I believe that medical physicists should not be directly involved in the
development and implementation of an integrated hospital information system but should be
active participants in this environment. They should, however, collaborate with other healthcare
professionals serviced by the integrated HIS in order to refine their occupational proficiency and
meet the challenging applications of new technical advances.
REFERENCES
1. R. Haux, "Health care in the information society: What should be the role of Medical
Informatics?," Methods Inf. Med. 41, 31–35 (2002).
2. P. L. Reichertz, "Hospital information systems—Past, present, future," Int. J. Med. Inf.
75, 282–299 (2006).
3. R. Lenz and K. A. Kuhn, "Towards a continuous evolution and adaptation of information
systems in healthcare," Int. J. Med. Inf. 73, 75–89 (2004). C. J. Atkinson and V. J. Peel,
420
"Transforming a hospital through growing, not building, an electronic patient record system,"
Methods Inf. Med. 37, 285–293 (1998).
4. K. A. Kuhn and D. A. Giuse, "From hospital information systems to health information
systems. Problems, challenges, perspectives," Methods Inf. Med. 40, 275–287 (2001).
5. C. M. Tempany and B. J. McNeil, "Advances in biomedical imaging," J. Am. Med.
Assoc. 285, 562–567 (2001).
6. M. Thakur and B. C. Lentle, "Report of a summit on molecular imaging," Am. J.
Roentgenol. 186, 297–299 (2006).
7. J. Knight, "Bridging the culture gap," Nature (London) 419, 244–246 (2002).
8. AAPM Report Number 79: Academic Program Recommendations for Graduate Degrees
in Medical Physics, Education and Training of Medical Physics Committee (American
Association of Physicists in Medicine, College Park, MD, 2002).
9. A. B. Wolbarst and W. R. Hendee, "Evolving and experimental technologies in medical
imaging," Radiology 238, 16–39 (2006).
10. R. A. Greenes and E. H. Shortliffe, "Medical informatics: An emerging academic
discipline and institutional priority," J. Am. Med. Assoc. 263, 1114–1120 (1990).
11. R. Haux, "Health care in the information society: What should be the role of medical
informatics?," Methods Inf. Med. 41, 31–35 (2002).
12. H. K. Huang, "Medical imaging informatics research and development trends—An
editorial," Comput. Med. Imaging Graph. 29, 91–93 (2005).
13. M. Dennis, M. Rzeszotarski, and W. R. Hendee, "To prepare radiology residents properly
for the future, their physics education should be expanded in breadth and depth, and should be
more quantitative and mathematically-based," Med. Phys. 30, 1955–1957 (2003).
421
10.4. There is no legitimate role for an applications service provider

in radiology
Paul G. Nagy and Charles E. Willis

OVERVIEW
Image management is becoming a major challenge in radiology as the specialty migrates to an

all-digital format for recording, retrieving and networking images. Some radiology leaders
believe that image management is too great a technical challenge to be handled internally, and
that contracting with an external applications service provider is the best way to proceed. Others
feel strongly that image management is key to effective radiology services, and that quality can
be ensured only by internal management of radiology images. This controversy is the topic of
Arguing for the Proposition is Dr. Paul Nagy. Dr. Nagy is a medical physicist at the Medical
College of Wisconsin whose expertise is in imaging science and information technology. He is a
recognized voice in the PACS community advocating the third generation of PACS where the
integration of information systems focuses on the clinical workflow of Radiology. His research
includes developing the next generation of storage technology through storage area networks,
distributed file systems, and load balancing clusters. He has developed open-source applications
to monitor the performance of a PACS. He has been an invited speaker on Integrated Healthcare
Enterprise at the annual meetings of SCAR, AHRA and RSNA.
Arguing against the Proposition is Dr. Charles Willis. Dr. Willis is Associate Professor in the
Department of Radiology at Baylor College of Medicine and a member of the Medical Staff of
Texas Children's Hospital in Houston, Texas. He is a diplomate of the American Board of
Radiology in Diagnostic Radiologic Physics and licensed to practice Diagnostic Radiological
Physics by the State of Texas. He serves as the Chief, Section of Electronic Imaging for Texas
Children's Hospital. He is a member of AAPM Task Group 10 for Computed Radiography and
Task Group 18 for Electronic Display Devices. He has authored numerous publications on
practical aspects of Computed Radiography and PACS.
FOR THE PROPOSITION: Paul G. Nagy, Ph.D.
Opening Statement
The business case for the remote application service provider (ASP) model is a reaction to the
market's slow adoption of picture archival and communications system (PACS). However a truly
competitive third generation of PACS has emerged that employs inexpensive off-the-shelf
hardware. This PACS generation uses robust software based on DICOM standards, and focuses
on radiology workflow. The remote ASP model does not provide economy-of-scale cost savings
compared with a third generation PACS, and adds performance bottlenecks as well as
422
vulnerabilities by running services from a remote location. The total cost of ownership for an off-
the-shelf PACS is estimated to be between $4 and $8 an exam, whereas the cost of an ASP PACS
is between $10 and $14 an exam. Radiology departments do not typically have in-house
information technology leadership to articulate the benefits and risks of different ways to
implement PACS. I believe that this is a role ideally suited for the diagnostic medical physicist.
PACS has a large appetite for storage. Fortunately, storage is also the fastest-moving segment of
the computer industry, even surpassing Moore's Law. The areal bit density of magnetic hard
drives has grown by 60% each year since 1987, and by 100% each year since 1995. From 1988 to
about 1994, PACS was the driving force behind the entire storage market, and we paid bleeding
edge prices for it. Today highly rack-dense multiterabyte storage-area networks can achieve a
storage solution at a cost of $10–20 per Gigabyte. The in-house cost of storing a typical CT study
(100 MB) on RAID is $1, and on tape is less than $0.25. There is no economy of scale by going
to a data warehouse offsite for any department generating more than 40–50 k procedures per
year.
Another concern with the ASP model is the security and performance of the network
transmission line connecting your facility with an ASP data center. In disaster recovery, transfer
rates are insufficient across a wide area network. Even with an expensive OC-3 high-speed 155
MBit network connection, the time to restore one terabyte of data is just under two days! Local
area networks can now run at gigabit speeds on existing CAT 5 cabling at one-third to one-fifth
the cost of sending data over a wide area network and restore the same data in two hours.
Lastly, ASPs have a big hook attached; they have your data. Without stringent stipulations in a
contract, you may have to buy back your own data in order to migrate to another PACS solution.
Moving 10 s of terabytes of data accrued over a few years by a reasonably sized institution, could
be cost prohibitive as well as take weeks to months of dedicated network transmission time to get
back.
Most diagnostic medical physicists coming out of programs today have the necessary computer
skills to understand, implement, and support a PACS in a hospital environment. There is a real
need for PACS professionals in today's modern radiology department. This need presents a real
opportunity for medical physicists who understand image quality and information technology to
make a significant impact on patient care.
Rebuttal
Dr. Willis is correct in his assertion that an "on-site" ASP model is just a business decision based
upon the financial model the hospital prefers. The term ASP is ubiquitous and overloaded,
making comparison confusing. I would like to suggest using the term "fee for use" as opposed to
"on-site" ASP. The "fee for use" model has the identical technology and architecture as a capital
PACS and as such I have no problems with it. If a hospital has difficulty procuring the capital
necessary, the "fee for use" model is a legitimate way to go.
The remote ASP PACS is a difficult sell compared with other industries where the ASP model
has been successful. Modality and workflow integration requires considerable customized work
that needs to be done onsite. The file size of radiographic data is also atypical of the ASP
industry. In the banking industry, the transaction size is on the order of kilobytes (KB) whereas
in radiology that transaction size is easily 1000 times larger. This larger data size drives the
transmission costs up over a wide area network at even minimally acceptable performance rates.
423
The ASP model has validity when a facility is too small to support the salary of an administrator.
That threshold is well under 50 000 procedures per year and can be calculated by looking at the
total cost of ownership between the two models. But I believe it is a Faustian bargain to rely upon
your ASP vendor for IT expertise. Computer technology is a disruptive industry and vendors do
not always pass on the benefit of lower cost technology. The physicist, who frequently acts as the
technology officer of the department, could justify the cost at even a modest sized institution.
It is very desirable for a hospital to have a physicist who can understand the entire imaging chain,
from image generation to image transfer to image review. Piggy backing PACS responsibilities
onto the physicist's role is a great opportunity for recent graduates looking for a way to break into
the industry and make a difference. The salary of a physicist can be easily cost justified compared
with the cost of not being an educated consumer in the PACS industry.
AGAINST THE PROPOSITION: Charles E. Willis, Ph.D.
Opening Statement
An ASP is a business model used extensively in banking and other industries to obtain up-to-the
minute information technology (IT) without investing in capital equipment and skilled IT
personnel. The application service provider assumes the costs of hardware, software,
maintenance, personnel and connectivity, and provides services that are critical to a customer's
business. This model has been proposed as a way for hospitals to obtain a PACS without making
a significant initial investment.1,2 The fee for an ASP PACS service would appear in the
institution's operational budget as an incremental cost of doing business, instead of as a
combination of capital budget items and operating expenses. In many ways, the ASP model
resembles leasing, which is used widely to obtain medical imaging equipment.
The notion that an ASP has no legitimate role in radiology reflects the idea that an institution
must retain control over medical images. However, most radiology operations outsource their
long-term archives of images offsite. Few departments have sufficient space onsite to maintain
even two years of films. The expanding availability of electronic imaging and low-cost, high-
volume storage paves the way for a return to the ideal of retaining image archives entirely within
the institution. This return would require an expanded radiology IT staff, which would be a
challenge considering the current demand for skilled personnel in the IT industry.
An ASP does not have to provide total IT support for a radiology department. It can play a more
limited role, such as supplying the radiology information system (RIS) while the institution
maintains the PACS system. An ASP could supply an offsite archive that serves as long-term
image storage or as a disaster recovery system. The ASP could staff and operate the complete
electronic imaging system, at the direction of the radiology department on the premises of the
hospital. This would be an example of an "on-site ASP."
Some early adopters of PACS are now facing substantial investments to avoid technological
obsolescence, in the sense that equipment and software cannot be reliably maintained. The ASP
model can be considered as a way for hospitals to avoid technological obsolescence, because the
responsibility for maintaining the IT system is transferred to the ASP. The operating cost for the
system is negotiated and can be budgeted appropriately.
424
There are many community hospitals that do not enjoy the capital resources to purchase IT
systems for radiology. They also lack the personnel to operate such systems. The ASP model is
often an attractive option for these smaller hospitals.
Demonstration of an ASP as a viable business choice depends on several factors. The first is
whether ASP can be cost effective. For an off-site ASP, broad–band connectivity to the hospital
is crucial. Frequently, hospitals that would find ASP attractive are underserved in terms of
network infrastructure. The specific terms of an ASP contract are especially critical, including
performance, reliability, and exit clauses. The hospital is responsible for the quality and
reliability of imaging operations: Whether an ASP can meet or exceed what the radiology
department can provide on its own must be carefully examined.
Rebuttal
Dr. Nagy claims that an offsite ASP will cost more to operate than will a "third generation, off-
the-shelf PACS." Because of numerous hidden costs, I remain skeptical of all PACS cost
estimates, but recent studies3 suggest that fee-for-use leasing has the greatest potential for cost-
effectiveness in smaller hospitals. Who is going to operate, train, maintain hardware, software,
and supplies for an "off-the-shelf PACS" system? Who is going to assure that the in-house
archive is still functional when its end-of-life occurs before statutory requirements for retaining
images are reached? Off-the-shelf systems eventually become obsolete and unsupportable, so
replacement costs should be considered.
Dr. Nagy's analysis of bandwidth limitations from an offsite archive is flawed. To maintain
continuity of clinical operations, it is only necessary to retrieve relevant prior images as needed,
and not to immediately retrieve the entire archive. The archive could be completely restored by
physically transporting magnetic media, rather than by "dedicated network transmission."
While I agree that most diagnostic physicists have the computer skills to understand PACS, I
strongly disagree that implementing and supporting PACS is an appropriate role. 4 While the
physicist can act as a valuable consultant to the hospital, this should not include performing
economic analyses for hospital administration or managing an electronic image distribution
system. "Medical physicists must utilize their training and expertise in the physics of imaging to
continue to provide the highest quality diagnostic information to the radiologist with the least
possible radiation or other safety risk to the patient, staff, or public." 5
REFERENCES
1. K. Dreyer, ASP Progress Report, Imaging Economics (Spring 2001), p. 46.

2. M. S. Marchese, "Image archiving the ASP way," Med. Imaging 15, 58–64 (2000).
3. B. Reiner, E. Seigel, and P. McKay, "Adoption of alternative financing strategies to
increase the diffusion of picture archiving and communications systems into the radiology
marketplace," J. Digit. Imaging. 13, 108–113 (2000).
4. P. Sprawls, "Characteristics of medical images for effective diagnosis: The expanding
role of the medical physicist," in The Expanding Role of the Medical Physicist in Diagnostic
Imaging, Proceedings of the 1997 AAPM Summer School (Advanced Medical Publishing,
Madison, WI, 1997), pp. 1–16.
5. C. E. Willis and S. K. Thompson, "Strategic moves: Medical physics in the digital
domain," Advance Admin. Radiol. Radiat. Oncol. 11, 30–35 (2001).
425
10.5. The provision of consultative physics services as commodities

undermines the professionalism of medical physics
Robert Kriz and Lincoln B. Hubbard

OVERVIEW
Medical physicists have traditionally been full-time employees of hospitals or healthcare organizations.
In this capacity, physicists have contributed to professional needs such as staff training, equipment
acquisition, policy development and procedure evaluation, in addition to technical concerns such as
equipment calibration, quality control and radiation safety. Increasingly, these technical concerns are
outsourced to consultants who offer physics services as commodities on an as-needed basis. This
approach decreases opportunities for physicists as full-time employees of institutions, and undermines
the professionalism of medical physics.
Arguing for the proposition is Robert J. Kriz, M.S. Mr. Kriz received his M.S. from DePaul University in
1976, and recently retired from the University of Illinois Hospital and Clinics in Chicago. He spent the
last 20 years working in diagnostic radiology writing purchase specifications, doing bid evaluations and
acceptance testing of radiological imaging equipment. He also developed and ran the physics and
radiobiology training program for radiology residents at the University of Illinois, and taught the
diagnostic physics course for over 20 years. He still teaches diagnostic radiological physics at Rush
University.
Arguing against the Proposition is Lincoln B. Hubbard, Ph.D. Dr. Hubbard is a radiological
physics consultant in the Midwest. A physics B.S. at the University of New Hampshire preceded
a Ph.D. in theoretical nuclear physics from M.I.T. A postdoctoral at Argonne National Lab under
Warren Sinclair was followed by 6 years of college teaching along with research at Oak Ridge
National Lab. In 1974 Dr. Hubbard changed from academic to clinical work, and he joined Fields
and Griffith in their consulting group the following year. He is certified by both the ABR and the
ABHP, and has been active with the ABR since the 1980s as an oral examiner and as a
contributor to the physics written exams. In the 1980s he assisted Dr. Lanzl in creating the
medical physics teaching program at Rush University. Also in the 1980s he was the lead physicist
in the Illinois Radiological Society's mammography accreditation program which became, in
1987, the ACR MAP program. He and Nancy, his wife of 41 years, have two daughters and two
granddaughters.
FOR THE PROPOSITION: Robert J. Kriz, M.S.
Opening Statement
Several trends that began some time ago are reducing medical physics to a commodity service
and seriously eroding the profession. A physics report is a commodity when its accuracy and
relevance cannot be judged by the administrator paying for it (e.g., a safety report on an x-ray
426
facility). However, if the state accepts the report, it is considered adequate. This can be especially
problematic if the agency is more likely to accept the report if written by a former state employee
who is now in the private sector.
I am chagrined at the diminishing number of diagnostic physicist positions in medium to large

hospitals. Where such positions do exist they tend to be in academic institutions dedicated to
research. Research is important, but it doesn't address the common problems of a large radiology
department. The issue is that the diagnostic physicist does not generate revenue, and therefore is
expendable so long as regulatory requirements are met.
Writing performance specifications and acceptance tests into bid documents is happening less
frequently, partly because of the formation of large purchasing groups. At my former institution,
there were several large purchases made without any physics input at all. In one case several
units were combined into a single purchase to force vendors to be more competitive. Because of
the size of the purchase, university management took charge of the negotiations, with disastrous
consequences. Price was the only consideration in these negotiations, and systems were
downgraded, needed options were dropped, less powerful generators were acquired, and low-end
imaging chains were substituted, all without the knowledge or advice of the physicist.
My state allows the bid procedure to be bypassed if the preferred vendor has contracted (with
large specified discounts) to a buying group. Often a vendor is preferred simply because the
selection reduces paper work and avoids complicated analyses of competing systems. In this
approach, physics advice is eliminated entirely. This is happening despite the advance of
technology and the increasing complexity of equipment. In these situations the physicist is
relegated to safety surveys, shielding calculations and distribution of film badges. Even in these
areas, especially surveys of x-ray equipment, the work is often transferred to nonphysics
personnel.
Safety surveys of x-ray equipment require strong (and formal) education in physics. Modern
R&F and especially special procedures/cath labs, have become enormously complex. Adjustable
filtration and automatic brightness stabilization that is contrast dependent is in use. A cookbook
inspector filling out forms, measuring HVLs and checking collimator accuracy may entirely miss
the point. But who educates the administrator? If the credentials and reports of the inspector are
accepted by the approval agency, and the financial investment is low, the administrator is happy.
Standards makers and regulators have also contributed to the problem. The Joint Commission on
Accreditation of Healthcare Organizations seldom reviews the credentials of physicists
performing equipment safety surveys. As a result, many of these surveys are performed by RTs
who purchase a dosimeter and fill out forms. Often the selection of physics support is made by a
chief technologist or business manager with more concern about cost than about expertise. We
need the help and support of our radiology colleagues to ensure that the standard of care is not
reduced. Many radiologists are not aware of this problem.
The FDA under MQSA has contributed to the problem by not insisting on high standards of
education, and by allowing the states to approve inspectors of mammography equipment. In
many cases states have decided that an RT with a short training course and some experience is
equivalent to an advanced degree in physics. This situation has improved somewhat recently but
needs much more improvement.
Rebuttal
427
I strongly agree with my colleague that the consulting medical physicist adds to the
professionalism of the discipline. But how is the decision made to go "outside," and who makes
it? In today's competitive and cost-cutting environment, a request for a consultant may be looked
upon as a shirking of responsibility. In actuality, it is more likely to be the opposite, especially as
technological sophistication increases. Can all hospitals perform IMRT, TIPS, digital
mammography, etc.? Competition demands they must. Costs may demand otherwise. Physics
support for these services is viewed by some as purely a cost without significant benefit (i.e.,
income). Radiologists understand these services cannot be performed without strong physics
support, but they are in their own financial battle.
Credentialing of medical physicists has become more important than ever. Statutory
credentialing may become necessary. Administrators are more likely to respect the credentialed
physicist as a professional. Once they are aware of Board Certification, for example,
administrators are less likely to accept reports from uncredentialed "look alikes."
When a regulatory agency accepts reports from less than fully qualified medical physicists, they
severely damage the profession.
AGAINST THE PROPOSITION: Lincoln B. Hubbard, Ph.D., FACR
Opening Statement
In 1902 there were no radiation physicists routinely working in the clinical environment. The
physician radiologist was a jack-of-all-trades who could manufacture photographic plates,
generate high voltage, get a gas x-ray tube to conduct, position a modest patient, and interpret
results. Fifty years later, radiologists had specialized assistants—including those jack-of-all-trade
medical physicists who are the colossi of our professional heritage.
Specialization is a major contributor to the "professional success" of radiology, both for the
"professionals" (radiologists, physicists & technologists) and for the population served.
Specialization occurs in many fields; in economics it is viewed as one of the wealth producing
engines of modern society. Specialization in medical physics is no exception. The diverse
offering of modern medical physics requires too may skills for a single jack-of-all-trades to be a
professional expert in all areas.
With a fragmented and highly specialized field, how are all of the special offerings to be
covered? The jack-of-all-trades cliché includes "master of none." That is, a full-time physicist
who limits medical physics services at an institution to his/her expertise is probably acting in an
unprofessional manner. On the other hand, a full-time physicist who performs services beyond
his/her professional capability is certainly unprofessional. The full-time physicist should
encourage the most appropriate medical physics services for his/her facility. If additional
expertise or effort is needed, then it should be provided by either additional full-time physicists or
by consultants. The choice between the two depends on efficiency. Does the use of a consultant
negatively impact a full-time medical physicists? Certainly not, if the use is appropriate and
properly integrated into the physics services of the facility. For example, infrequent tasks such as
shielding design, beam data acquisition, or acceptance testing might be beyond the physicist's
expertise, or involve too much effort to be handled by the regular staff. Certain routine tasks may
also be better handled by consultants than by full-time staff, due to time, interest or expertise
limitations.
428
Mammography physics is one area where physics professionalism has grown rapidly in the last
decade. Here specialization with extensive first-hand experience is recognized as essential both
within the clinical area and by regulatory and accrediting bodies.
There is nothing unusual about reducing a highly specialized professional service to a

commodity; this is simply fee-for-service. The professional paragons: medicine, law and
accounting, make widespread use of fee-for-service. Services delivered without fee-for-service,
i.e., only by employees, such as military officers and bureaucrats, are usually considered less
"professional." When fee-for-services exists, so do consultants. It is the consultants who set the
professional standards and fees. It is the market activity of consultants that allows a balance
between a client's needs and the professional's service. Economics tells us that goods and services
produced under conditions far removed from the free-market situation are inferior in the long
run. Medical physicists whose standards and/or fees are set without significant marketplace input
deliver inferior (and thus, intrinsically unprofessional) value. If medical physicists as a group
manage to follow such a course, they will achieve an inferior professional status in the long run.
Rebuttal
My colleague is correct in stating that a problem with physics professionalism is the presence of
self-proclaimed physicists who know little about actual physics, and yet act as consultants. But
my colleague errs in suggesting that eliminating consultants will eliminate the problem. As with
most multifaceted endeavors, some routine tasks carried out by legitimate physicists could be
handled by technicians with less training. As with nurse-physician or paralegal-attorney
relationships, these technicians need adequate supervision to ensure that conditions are not
mishandled because they are misunderstood. It is the independent and unsupervised activities of
the inadequately trained that usually lead to services of poor quality.
To a considerable extent, well-trained physicists are to blame for this situation. We have
burdened ourselves with protocols designed to make our tasks straightforward. To the
uneducated, these protocols make the tasks look easy. Often we end up with an approach that
mandates one method, even though alternate methods may yield satisfactory results. In many
imaging evaluations, for example, four repeat measurements are required to determine a
coefficient of variation. This number is not derived from statistics, but is a purely operational
technique to allow the statistically incompetent to make valid "judgments."
My colleague is also correct in his other major point that medical physicists are often
underutilized. He identifies part of the cause of underutilization as the commodity nature of
consulting. The remainder of the cause can be attributed to misguided or under applied standards
of accrediting and regulatory bodies. I agree with the comment about underutilization, but believe
that the cause of this problem does not lie primarily with organizations such as the JCAHO or
ACR. We medical physicists have not defined the core services that we have a unique capability
to perform. Until we define these services, and market ourselves as being defined by them, they
won't sell very well to organizations whose interest in medical physicists is less consuming than
our own. AAPM committees spend too much time with the promulgation of protocols and too
little time defining why and how we can make clinically important and cost-effective
contributions to health care. And almost no time is spent developing arguments to describe why
our training and experience gives us the legitimacy to provide these contributions. If we really
want a profession, we must define it, legitimate it and sell it.
429
10.6. The growth of biomedical engineering is a major challenge to

medical physics
Randell L. Kruger and Bruce H. Curran

OVERVIEW
Educational programs in biomedical engineering are rapidly establishing and growing, in large
measure because of funding from the Whitaker Foundation and National Science Foundation. In
these programs, the most popular instructional track is imaging. Some diagnostic physicists feel
that this pipeline of imaging-trained biomedical engineers is a major challenge to physicists in
imaging. Others think that this influx of engineers is an opportunity that should be capitalized on.
This difference in perspective is the subject of this month's Point/Counterpoint.
Arguing for the Proposition is Randell Kruger, Ph.D. Dr. Kruger is the Medical Physics Section
Head in the Radiology Department of the Marshfield Clinic. Dr. Kruger received his Ph.D. from
the Medical College of Ohio and completed a post-doctoral medical physics residency at the
Mayo Clinic. Prior to his doctoral program he earned a master's degree in mechanical engineering
from Arizona State University. He has seven years of engineering work experience with the U.S.
Air Force and Allied-Signal, Inc. He is certified in Diagnostic Physics by the ABR and is
president of the North Central Chapter of the AAPM.
Arguing against the proposition is Bruce Curran, ME, MS. Mr. Curran received his Masters
Degrees from Dartmouth College (Engineering Science-Biomedical Engineering) and
Northeastern University (Computer Science). He is Clinical Assistant Professor of Radiation
Oncology at the University of Michigan and responsible for clinical physics within the
Department of Radiation Oncology. He currently serves as chair of the Meeting Coordination
Committee of the AAPM and co-chair of a task group on clinical implementation of Monte Carlo
dose calculations. He is a fellow of the AAPM and the ACMP.
FOR THE PROPOSITION: Randell L. Kruger, Ph.D.
Opening Statement
Can an engineer become a medical physicist? I am a personal testimonial that engineers can and
do migrate into medical physics, after receiving the proper educational and clinical training.
During the 2003 AAPM Annual Business Meeting in San Diego the topic of changing the
academic requirements for AAPM membership was discussed. The proposed amendment adds
two words to ARTICLE IV, Section 4 of the Bylaws—they are ("or Engineering" added to the
existing text of Physical Science). This change would add engineering degrees to the criteria for
AAPM Membership eligibility. The motivation for the change is the need to create consistency
between current practice and the bylaws. However, some diagnostic medical physicists are
concerned that imaging-trained biomedical engineers would challenge the role of, and seek to
replace, the diagnostic medical physicist.
430
The clinical and research applications of medical imaging in bioengineering have contributed to
the explosive growth of biomedical engineering jobs. 1,2,3 Of the more than 100 college and
university programs that offer academic programs in biomedical engineering, more than half
offer imaging educational or directed-research programs.1 Significant job growth and interest in
biomedical imaging has been accelerated with the lure that "all teaching hospitals, have a
growing need for bioengineers trained in imaging methods." 2 The U.S. Labor Department's
Bureau of Labor Statistics projects that the number of biomedical engineering jobs will increase
by 31.4 percent through 2010.1 Are all of these imaging-trained biomedical engineers planning to
work for industry or in research? The National Institutes of Health Bioengineering Consortium
provides a definition of bioengineering, which does not include the word "imaging" anywhere in
its 59-word statement.4 Yet the rapid development of a biomedical imaging curriculum and
career field in biomedical engineering indicates a shift in focus of the biomedical community.
The roles of the medical physicist in diagnostic imaging have been well documented and
comprehensively defined by the AAPM, the American College of Radiology (ACR), and the
European Federation of Organisations for Medical Physics.5,6,7 These organizations have
described and defined the diagnostic medical physics professional role, and the practice, training,
and qualification requirements in the field. A primary responsibility of the diagnostic medical
physicist is the development and supervision of a quantitative quality control program. However,
the diagnostic medical physicist has several other responsibilities and duties (such as: radiation
safety; compliance activities; radiobiological, shielding and equipment evaluations; educational
activities; and research, to name just a few). An imaging-trained biomedical engineer is not
prepared or trained to perform these duties and responsibilities. Most members of the biomedical
engineering and medical physics communities understand the differences between a diagnostic
medical physicist and a biomedical engineer. The concern is that other members of the medical
community might assume (or be misled to understand) that an imaging-trained biomedical
engineer can perform the duties and responsibilities of a diagnostic medical physicist. This would
jeopardize the quality of diagnostic imaging services provided to the medical facility and its
patients.
Rebuttal
I agree with my colleague that medical physics is an applied branch of physics that deals with the
application of physical principles to the diagnosis and treatment of human disease.8 However, I
disagree with his statements that link engineers and medical physicists. The logic he employs to
support the equivalence of biomedical engineering and medical physics is flawed.
Medical physics is a focused field of study that requires clinical training or preceptorship.
Biomedical engineering is a broad interdisciplinary field of study with little or no clinical
training. A description of biomedical engineering provided from a large state university
biomedical engineering department9 states "the Biomedical Engineering Graduate Program is an
interdisciplinary program designed to provide broad familiarity with the interactions among the
engineering, biological and medical sciences and in-depth training in one of the traditional
engineering disciplines." Medicine in general is an application of science to the treatment of
human disease and health, and its practitioners are educated and trained specifically for expertise
in their field. It appears my colleague proposes an exception to this rule for biomedical engineers.
Medical physics is significantly influenced by the technological advances, as is all of medicine.
An individual with broad familiarity would lack the specific training and experience necessary to
provide the required clinical services.
431
I think it is important to consider the fundamental factors driving this issue. The Whitaker
Foundation's funding has significantly accelerated and expanded educational programs in
biomedical engineering. The expansion of biomedical engineering into medical imaging,
interestingly, comes at a time when the medical physics profession is experiencing a shortage of
practitioners and a limited number of training programs. Donald Frey's statement 10 "one of the
more serious problems facing the profession of medical physics is the shortage of practitioners"
highlights this problem. The laws of supply and demand cannot be ignored.
Can an engineer become a medical physicist? The answer is yes, provided he or she obtains the
proper academic preparation and clinical experience.
AGAINST THE PROPOSITION: Bruce Curran, ME, MS
Opening Statement
According to the AAPM, medical physics is "an applied branch of physics concerned with the
application of the concepts and methods of physics to the diagnosis and treatment of human
disease." 11 This definition focuses on the application of training and experience to the diagnosis
and treatment of patients. There are few (if any) medical physicists engaged in pure research
without thought to its future implementation, which distinguishes us from many of our
colleagues engaged in more theoretical branches of physics (defined, at least from one source, as
"the science of matter and energy and of interactions between the two, ").12 An interesting
observation on these definitions is that, for many universities, education in the field of "Applied
Physics" often appears under the domain of the College of Engineering.13
Appropriate to this discussion is a look at the profession of engineer. One dictionary defines an
engineer as "one who is trained or professionally engaged in a branch of [engineering] the
application of scientific and mathematical principles to practical ends such as the design,
manufacture, and operation of efficient and economical structures, machines, processes, and
systems." 12 Since physics is clearly a member of the sciences, it appears that engineers are
individuals who can also be considered to be involved in the application of physics to the
solution of a certain class of problems such as the diagnosis and treatment of human disease. It
would thus seem that, with a slight twist on the origins of the phrase, "We have met the enemy
and he is us." 14
For the majority of medical physicists today, technological advancements in imaging and therapy
have led to a new role for the medical physicist, namely that of manager of the complex
equipment necessary to our profession. We are no longer expected only to understand how
different radiations interact with materials and patients. Today, physicists must also be
knowledgeable about computer systems, networks, and the myriad of new technologies essential
to current clinical practice. The influx and influence of individuals with advanced training that
includes an in-depth understanding of the technology itself is helpful, perhaps even necessary, to
effectively carrying out our duties, as well as advancing state-of-the-art patient care. A
collaborative environment that includes professionals with skills both in physics and engineering
appears to be the best of all worlds.
Patients benefit from having a team of individuals with a broad range of skills available for
designing, building, testing, and monitoring the techniques and equipment needed in the practice
of medical physics. These skills require significant education, training and experience, and it is
432
unlikely that any single individual will master all aspects. The inclusion of biomedical engineers,
with their strengths in equipment and biological/equipment interfaces, in the profession of
medical physics will strengthen our profession and allow it to grow. This in turn will improve our
stature and acknowledgement as key individuals in the diagnosis and treatment of patients.
Rebuttal
One might as well ask "Can a theoretical nuclear physicist become a medical physicist?" The
answer of course, is yes, as many of our colleagues can attest. Did their initial education
completely prepare them for our field? Probably not. As Dr. Kruger notes, proper education and
clinical training is necessary for most individuals entering our field, whatever their educational
background.
Does an education in biomedical engineering prepare individuals less well for entering our field?
It certainly prepares them differently. A biomedical engineer specializing in biomechanics would
be no more suitable for clinical practice than the theoretical nuclear physicist. A review of the
course offerings in a biomedical engineering program reveals courses in anatomy,
instrumentation, physiology, radiological health, imaging (radiation, MR, optical), and medical
imaging systems,15 all appropriate to our profession.
So how do we "separate" those engineers (and physicists) not appropriately qualified to practice
medical physics from those who are? Ideally, the certification/licensure process would ensure
that only qualified individuals attain the title of medical physicist. The reality is, however, that
many individuals are given the title long before they acquire the skills necessary for practice.
This is mostly a result of history; the small number of educational programs in medical physics,
the lack of appropriate residence and training programs that give us the time to acquire needed
skills before certification, and the rapid increase in the need for properly trained professionals in
our profession.
Medical physics as a career will continue to attract a polyglot of engineering and scientific
professionals. It offers the alluring combination of interesting, challenging problems, the
satisfaction of helping humanity, and good salaries and benefits. The incorporation of such
diverse backgrounds has helped to keep the field fresh and innovative. We should continue to
encourage entry into medical physics of persons with diverse backgrounds, while striving to
improve the processes by which we identify those individuals who have earned the title of
medical physicist.
REFERENCES
1. www.whitaker.org
2. http://summit.whitaker.org/white/imaging.html
3. http://summit.whitaker.org/white/basic.html
4. www.becon.nih.gov/bioengineering_definition.htm
5. www.AAPM.org
6. www.ACR.org
7. The European Federation of Organisations for Medical Physics (EFOMP) Policy Statement
Number 5, "Departments of Medical Physics—advantages, organization and management," Physica
Medica XI(3), 126–128 (1995).
8. ACR Guide to Medical Physics Professional Practice, http://www.acr.org
9. www.whitaker.org/academic/database/index.html (search under: University of Minnesota)
433
10. January/February 2004 AAPM Newsletter

11. www.aapm.org/org/aapm_fact_sheet.html
12. The American Heritage® Dictionary of the English Language: Fourth Edition, 2000.
13. www.engin.umich.edu/departments/
14. Walt Kelly, Earth Day Poster, 1970.
15. http://www.engin.umich.edu/students/current/academics/courses/biomed.pdf
434
10.7. Through their preoccupation with new technical developments,

physicists have lost sight of the realities of cancer care and statistics.
Robert J. Schulz and James A. Deye

OVERVIEW
The history of radiation oncology is a marvelous story of technical developments leading to

improvements in the care of cancer patients. Many of these developments have increased the
precision of treatment and the (tumor)/(nontumor) dose ratio to yield higher cure rates with fewer
side effects of irradiation. There may be, however, a point beyond which the biological
variability of patients, rather than the imprecision of treatment, becomes the dominant influence
on treatment success. Whether we have reached, or gone beyond that point, is the subject of this
Arguing for the Proposition is Robert J. Schulz, Ph.D. Dr. Schulz is a Charter Member and
Fellow of the AAPM, Fellow of the American College of Radiology, and Diplomate of the
American Board of Radiology. He delivered the Earle Gregg Memorial Lecture and has twice
been a recipient of the Farrington Daniels Award. He was the Chairman of the Subcommittee on
Radiation Dosimetry (SCRAD) and of Task Group 21 for high-energy photon and electron
dosimetry. His professional career began at Memorial Sloan-Kettering (1952–1956), developed
further at the Albert Einstein College of Medicine (1956–1970), and concluded at Yale
University (1970–1992). His retirement to rural Vermont has enlarged his perspective on
radiation oncology.
Arguing against the Proposition is James A. Deye, Ph.D. Dr. Deye is Program Director and
Expert in medical physics with the National Cancer Institute, Radiation Research Programs. His
Ph.D. is from Vanderbilt University based upon nuclear physics research completed at Oak Ridge
National Laboratory. He has been a member of numerous committees within the AAPM, ACR,
and ACMP and chaired the AAPM Professional Council in the mid 1980s, where he founded and
hosted the first Tri-lateral Committee meeting. He is a Fellow of the AAPM and board certified
in radiation oncology physics by the ABMP and in radiological physics by the ABR.
FOR THE PROPOSITION: Robert Schulz, Ph.D.
Opening Statement
Although originally applied to stock brokers, the term "irrational exuberance" is also appropriate
for some medical physicists. By suggesting that IMRT is the holy grail of radiation therapy, it
appears that these physicists have lost sight of the realities of cancer and its management.
Although highly profitable for manufacturers and good for the employment of physicists, it is
unlikely that IMRT will reduce overall cancer mortality.
The rationale for IMRT is simple: a) the achievement of local control will reduce mortality; b)
the likelihood that this achievement increases in some unspecified way with tumor dose.
435
Although local control is likely to reduce local recurrence, about 70% of cancer deaths are due to
metastatic disease despite aggressive treatment of the primary. This is clearly the case for breast
cancer where local control is readily achieved but mortality has not changed in 70 years. And in
cancers in which death is caused by failure to achieve local control, e.g., brain and ovarian
tumors, there is little reason to expect better outcomes by simply increasing the dose. As with
surgery, the goal of IMRT is to "resect with negative margins" while avoiding the trauma and
mortality of operative procedures. However, surgical resection fails for many cancers. Why
should we expect more from IMRT? Surgical resection with adequate margins is standard
therapy for esophageal cancer, for example, but five-year survival remains a dismal 12%. The
achievement of local control has been a mantra for cancer therapy for many years, but relatively
static mortalities suggest that newer, more sophisticated, systemic therapies will have to be
developed before significant improvements are obtained.
In the U.S. this year, cancers at twenty specific sites will account for over 93% of the estimated
1.27 million new cases and 91% of the 553 thousand deaths. Surgery is the primary therapy at
eleven of these sites and chemotherapy at three, with radiation primary or competitive with
surgery at only six. Excluding skin, the remaining five (uterine cervix, endometrium, oral cavity,
brain and prostate) account for 24% of incidence and 12% of mortality (with the prostate
responsible for 15.6% and 5.7% respectively). As IMRT is mainly envisaged as replacing or
supplementing conventional radiation treatments, it is unlikely to change the distribution of
therapies, so what gains may be realized by IMRT will have to be made at those five sites. If, and
this is a very big "if," IMRT could reduce the 12% level of mortality by one tenth, and assuming
that it is employed for all patients in four of the five sites and for half of all prostates, then the
overall number of deaths would be reduced by about 4900. To realize this gain, each of the
approximately one thousand radiation centers in the U.S. would need one IMRT system. Each of
these systems would at best extend the lives of five patients per year. How can this huge capital
investment and operational expense be justified?
Rebuttal
It appears that Dr. Deye may be amongst those who "have lost sight of the realities of cancer care
and statistics." At no point in his argument does he discuss even one form of cancer. He eschews
statistics and mortality trends, and presents little evidence to support the introduction of complex
and expensive technologies such as IMRT. Does Dr. Deye endorse the idea that "the marketplace
will judge the cost/benefit ratio" for a new technology like IMRT? Certainly he knows that
expenditures for radiation therapy are steadily increasing while reductions in cancer mortality
over the past 20 years are at best modest and probably unrelated to improved radiation delivery
systems.
Although not large, the number of patients who have received IMRT is sufficient to provide a
reasonable assessment of its potential benefits. A recent Medline search using the search term
"IMRT and prostate cancer" identified 30 citations, none of which showed IMRT to have any
impact (positive or negative) on morbidity or mortality. As this "darling" of IMRT can be as
effectively treated by conventional radiation, implants or surgery, one wonders why virtually
every current and future owner of an IMRT system offers prostate treatment as the principal
justification for its purchase.
Some of us recall earlier efforts to improve radiation therapy such as 70 MV x rays, hyperbaric
oxygen tanks, neutron beams, and negative pi mesons. At the time, the rationale for each of these
made as much sense as IMRT does today. However, whereas these earlier experiments were tried
436
on a limited scale and at relatively low cost, IMRT is market driven (as the commercial exhibits
at AAPM and ASTRO meetings clearly demonstrate). It appears that competition between
hospitals for patients plays a far larger role in purchase decisions than do expectations of
improved clinical results.
AGAINST THE PROPOSITION: James Deye, Ph.D.
Opening Statement
This is not a new concern. It might have been voiced about physicists who looked for ways to use
radioisotopes to quantify organ function or morphology, accelerators to replace Cobalt 60 units,
and computerized treatment planning to expand point dose calculations, to name just a few. What
all such events have in common is the "preoccupation" of scientists, especially medical
physicists, with the utilization of recent scientific or engineering accomplishments for medical
purposes. In each case scientists were pursuing new avenues of diagnosis or treatment as opposed
to just performing functions in support of the status quo. And in each case it was not a given that
there would be a clinical benefit relative to that status quo (e.g., 4 MV linacs over Cobalt 60 and
electron beams over orthovoltage). One may wish to argue that some of the now-accepted
technologies have not truly advanced the care of cancer patients. If so, then let's weigh these
standard treatment and diagnostic methods on the basis of patient outcomes and eliminate or
curtail those that don't measure up. However, outcomes data cannot be used to predict which new
technologies will be beneficial, and we impede true advances if we decide on the basis of
personal guesses or biases. As always there needs to be a balance between the new developments
and the support of quality in standard practice. However, I am unaware of any data to suggest
that this balance has been lost.
The proposition may be construed to mean that some of these new tools are moving too quickly
into standard practice before their safety and efficacy have been adequately demonstrated. Yet,
even if true, this argument would say that physicists are not spending enough time on these new
developments rather than too much. In addition, if the proposition is extended to say that the
implementation of some of these new technologies is so labor intensive as to detract from
required clinical needs, this just underscores the need for clinical facilities to forego the use of
these new methods until they have the appropriate staffing. To do otherwise jeopardizes patients
and puts institutions at risk, but it should not be used as an indictment of the pursuit of the new
technologies. It may be that some physicists have abrogated their role in advising clinicians and
administrators on required staffing needs. And it may be that some clinical physicists are not
trained to the level required by these new techniques. Yet neither of these possibilities should
short-circuit promising developments.
There is ample reason to believe that many new technologies (from bio-imaging to IMRT) will in
fact improve cancer care. Whether this can be demonstrated with a probability sufficient to justify
the cost is a question that goes well beyond science, since it depends also on economics, politics
and societal needs. But as was said in the opening sentence, "this is not a new concern" nor is it
unique to medicine. It is the duty of the scientist to get the science right and to let those who deal
with the big picture put things into perspective. The marketplace will judge the cost/benefit ratio,
clinicians will judge the quality of patient care, and society will balance the agenda. It would be
arrogant of the physicist to preempt any of these roles.
Rebuttal
437
My esteemed colleague is using the debate technique known as "begging the question." The
question is not just IMRT or cancer mortality; it is technology in the whole cancer picture. In this
setting one must consider morbidity and quality of life issues in addition to mortality. Many
studies are proving that new image-guided therapies, of which IMRT is one, are allowing
increased doses to tumors with smaller doses to normal tissues. Zelefsky et al. observed one-third
the degree of rectal toxicity at 81 Gy to the prostate with IMRT compared with 3D CRT.1 This is
a major issue for patients who live a long time after treatment even if they are not "cured."
Reduced morbidities and greater patient convenience can make radiation therapy the treatment of
choice for tumors that historically may have had surgery as their primary treatment (e.g., breast
cancer, prostate, brain). Thus, an argument based simply on mortality, underestimates the true
value of IMRT as more sites become amenable to radiation therapy because of its increased
precision and understanding of treatment bio-effectiveness extending beyond the physical dose.
Similar considerations apply to other treatment technologies such as HDR, permanent seed
implants, proton therapy, radio-immunotherapy and many variations of IMRT such as
tomotherapy and robotic arms.
On the diagnostic side, fMRI, MRS and PET are delineating targets that require precision
treatment delivery. They are also providing new information that is affecting treatment decisions
and Quality of Life issues for incurable patients. These decisions are contributing to development
of a more comprehensive picture of the disease process and offering appropriate treatment
choices at different stages of disease.
Clearly there is a balance to be struck between the cost of new technologies and the probability of
benefit. This question goes well beyond the expertise of the physicist who is asked to explore the
true potential of many very exciting technologies. If physicists do their part with scientific rigor,
others will be able to use these results to fill in the rest of the picture. In this manner physicists,
along with many other interested parties, will be able to judge the art that rests on the science of
medicine.
REFERENCES
1. Zelefsky, presented at 6th International Symposium on 3D CRT and IMRT, June 29–July
1, 2001.
438
10.8. Medical physicists would benefit by establishment of an

Institute for Biomedical Imaging
Philip F. Judy and Stephen R. Thomas

OVERVIEW
Both the U.S. House of Representatives and the U.S. Senate are considering bills to establish a
National Institute of Biomedical Imaging (NIBI) within the National Institutes of Health. The
House bill (HR-1715), introduced in May, 1997 by Rep. Burr (R-NC) has been referred to the
House Committee on Commerce‘s Subcommittee on Health and Environment. The Senate
proposal (S-990), introduced by Sen. Faircloth (R-NC) in July, 1997, has been referred to the
Senate Committee on Labor and Human Resources. As stated in both bills, the purpose of the
NIBI is to ‗‗conduct and support research, training, the dissemination of health information, and
other programs with respect to radiologic and other imaging modalities, imaging techniques, and
imaging technologies with biomedical applications. As described by William Thompson MD,
Chairman of Radiology at the University of Minnesota in a recent issue of Medical Imaging,
‗‗It‘s critically important for us to try and get all the dollars centered in a biomedical imaging
institute, and to distribute those dollars in a much more effective way than we ever
have.‘‘ The NIBI would be guided by an Advisory Committee consisting of 6 scientists,
physicians, and health professionals knowledgeable about imaging, and 6 scientists, physicians,
and health professionals from other disciplines.
Arguing for the Proposition is Philip Judy. Philip F. Judy is Director of the Division of Physics
and Engineering and Associate Professor of Radiology at Brigham and Women‘s Hospital and
Harvard Medical School. His early research, which developed methods to measure image quality
of CT images, led to his present research that studies how variations in image quality affect
radiographic-image perception. He is studying how variations in breast parenchymal patterns,
genes, family, and race affect the perceptual and cognitive processes associated with the
detection of breast cancer using mammography. The goal of this research is to produce earlier
and more specific diagnoses from mammography. His past research has investigated lung and
liver nodule detection using CT imaging. The long term goal of his research is to develop
methods to reliably predict the clinical usefulness of novel imaging technologies for engineers
who are developing and optimizing image technologies, and for health care policy makers,
radiologists, and radiology administrators.
Arguing against the Proposition is Steve Thomas. Stephen R. Thomas is Professor and Director
of Medical Physics within the Department of Radiology, University of Cincinnati College of
Medicine. He has recently served as President of the American Association of Physicists in
Medicine (1997). Dr. Thomas earned his Ph.D. degree from Purdue University in 1973 in solid
state physics, and made the transition into medical physics through a postdoctoral fellowship at
the University of Cincinnati in 1974. His research interests include F-19 MRI of perfluorocarbon
compounds, and radiopharmaceutical dosimetry. He has served as a member of the NIH
Diagnostic Radiology Study Section (1993–1996).
439
For the proposition: Philip F. Judy
Opening Statement
The introduction of bills in Congress to create the National Institute of Biomedical Imaging
(NIBI) is the consequence of a 25-year effort by radiologists and medical imaging scientists. For
the past few years I have been involved in that effort as the AAPM representative on the Board
of Directors of the Academy of Radiology Research (ARR). The ARR is an organization of more
than 20 medical imaging societies that is promoting the creation of the NIBI. The organizing
principle of this effort is that the National Institutes of Health (NIH) are not structured to support
the investigation of fundamental medical imaging questions.
Disease or organ questions determine the structure of the NIH. Consequently, the fundamental
questions that emerge from NIH supported enterprises are addressed by molecular biology.
Important medical imaging questions are neither disease-specific nor organ-specific. The
answers to such questions emerge from physics, mathematics, computer science, and
psychology. The creation of the NIBI, if achieved, will benefit the medical physicists who would
receive increased grant support because their questions would emerge and their grants would be
funded.
Two questions remain. What is the likelihood that the NIBI will be created and, if it is created,
that it will deliver? The answers to the two questions are linked. Since the NIBI is promoted by
ARR, which is dominated by radiologists, why did they not propose the creation of the National
Institute of Radiology? To attract the necessary congressional support, the ARR leaders
understand that the new institute must have a wider constituency than just radiologists. One
constituency are the members of the 16 biomedical engineering organizations constituting the
American Institute for Medical and Biological Engineering (AIMBE). The AAPM has
representatives on the AIMBE Board and the current AIMBE president is a medical physicist
(W. Hendee). I believe that the creation of NIBI requires an alliance of ARR and AIMBE. An
effective alliance necessitates compromise, and that compromise naturally promotes the medical
physics agenda. In any case, the creation of the NIBI seems to be on track and, its charter will
increase the stature of medical physics within the NIH.
NIBI will have other roles that will provide opportunities for medical physicists to contribute to
improved national health. NIBI will coordinate medical imaging research both within the NIH
and in other federal agencies. One might argue that the chaos that exists in the funding of
medical imaging research is a result of the lack of an effective medical imaging voice at the
higher levels of decision making at NIH. NIBI would have authority to establish high level
contacts with industry. Medical physics imaging research falls in the cracks between federal
funding and industrial funding. A consensus regarding the appropriate roles of industry and
government in funding medical imaging research will increase opportunities for medical
physicists. NIBI will have a role as a third party with industry and regulators regarding the
appropriate methods of evaluation of imaging systems. Methods of image evaluation are an
important part of medical physics.
Rebuttal
Dr. Thomas argues that the NIH funding of biomedical imaging research is generous. Any
argument that a large or appropriate amount of the NIH funds are spent on biomedical-imaging
research is suspect until the biomedical-imaging expenditures are accurately classified.
440
Typically, the classification is done by officials at the NIH who are attempting to impress you
with the magnitude of resources that they are devoting to your problem. We must distinguish
between the cost of using mature imaging technologies and biomedical imaging research. One
consequence of the effort to create the NIBI is an increasing sophistication at the NIH of what
constitutes biomedical-imaging research.
Biomedical-imaging research is cut back disproportionately during funding downturns because

its cost is included in the cost of the use of imaging technologies. It is reasonable to reduce
purchases of expensive technology not critical to your mission when your budget is reduced or
does not grow as expected. The reason that entrenchment leads a disproportionate decrease in
funding of biomedical-imaging research is the perception at the NIH that biomedical-imaging
research is not a specific national health care mission. The goal of creating the NIBI is to change
that perception of decision-makers and avoid yo-yo funding of biomedical imaging research.
Dr. Thomas poses some specific and important questions that deserve more extensive debate
than is possible in this forum. The following are brief responses to these questions.
Some of the cost of imaging at the NIH is better classified as purchases of mature imaging
technologies rather than research. Monies would remain at the other institutes to purchase such
mature imaging technologies.
The probable additional administrative cost of the NIBI is a point well taken, but this problem is
not inherent to the proposal to create the NIBI, which has the potential to create savings by
reducing redundancy. It is, however, inherent to our democratic government.
The NIBI with an appropriate charter would reduce biomedical imaging turf conflicts rather than
increase them within the entire federal government not just the NIH.
Separation of medical imaging scientists and skilled image clinicians (radiologists) within the
medical school and hospital is counterproductive. As medical physicists we have learned that
close collaboration with physicians promotes technological innovativeness. Such close
collaboration is the paradigm that medical physicists have perfected, and we bring that
experience to these deliberations. I hope that increased funding of biomedical imaging would not
encourage deans of medical schools and presidents of hospitals to physically and
administratively separate basic imaging scientists from their clinical colleagues.
Just like other basic scientists, biomedical-imaging scientists will have to make a continuing
effort to inform the public and Congress about their contributions to national health. The answers
to the basic physics, computer science and psychology problems that emerge from biomedical
imaging are as directly applicable to specific medical problems as the answers to the basic
chemistry problems that emerge from the study of the human genome.
Against the proposition: Stephen R. Thomas
Opening Statement
As a medical physicist engaged in research and a former member of the Diagnostic Radiology
Study Section, I am intrinsically in favor of any development which would enhance productive
441
biomedical research. However, a devil‘s advocate role is appropriate as we define issues related
to the proposed NIBI and question whether major concerns have been addressed.
Let us examine the basic premise for establishing a new institute, namely, that under the current
organization by disease and organ system, the NIH is ill-equipped to support a discipline such as
imaging which extends across these boundaries. It is argued that there is no home for basic
investigations that promote the science of imaging as a primary mission. But are we convinced
that imaging has indeed been ill-served under the old roof? From 1995 to 1997, NIH funding to
academic radiology departments increased by 14% to $105 million, primarily channeled through
the NCI. A like amount from the NIH went into other imaging grants bringing the 1997 total to
approximately $200 million. Current 1999 federal budget projections include a $250 million 5-
year increase in NCI‘s $2.5 billion annual budget, and medical imaging as a specialty is expected
to receive one-fourth of that increase. Congressional bills specify that existing administrative
resources within the NIH will be utilized for NIBI and that appropriations for the new institute
will equal the amount currently obligated by the NIH for biomedical imaging. Whether or not
new funds would be identified is ambiguous at best. Under the scenario in which monies are
withdrawn from existing institutes, it would be difficult for those institutes to justify supporting
projects involving imaging. Ultimately, entrenchment may lead to an inability to fulfill specific
national healthcare missions.
The benefit of this institute to medical physicists should be defined in terms of its ability to
enrich the research environment. However, with a narrowed focus on imaging technology
potentially separated from the clinical setting, the outcome might be to exclude medical
physicists as the workload is assimilated by biomedical engineers and scientists/investigators
from non-medical backgrounds. Before establishing this new institute, fundamental questions
must be answered:
Through what formula will monies be designated and transferred from other institutes to
support the new institute?
What fraction of the transferred monies would be siphoned off for administration of the
new institute as opposed to productive research?
Will the public (and Congress) enthusiastically continue to direct funds toward
‗Radiology‘ rather than toward recognized and clearly tangible disease categories as
national killers (e.g., cancer, cardiac disease, etc.)?
Would the new institute be integrated effectively within the NIH to improve the health of
the nation, or would developing antagonisms (turf rage) toward this new entity prove
counterproductive to this mission?
Would innovative technological advances be impaired as a result of effective separation

between imaging science and clinical disciplines?
I am not convinced that answers to these questions have been formulated satisfactorily.
Certainly, the impact on the research medical physicist has not been unambiguously forecast.
The forte of the medical physicist is in the combined worlds of medical science and direct
clinical applications. We must be assured that imaging will not be isolated as a technical
discipline with disfranchisement of the practicing medical physicist.
442
Rebuttal
Phil Judy has indeed raised the critical issue—If created, will NIBI be able to deliver? There are
a number of forces assembled that might conspire to deny this opportunity. Some of them have
been alluded to in my opening statement. Consummation of an alliance between the ARR and
AIMBE represents only a single piece of the puzzle. The leadership at NIH must truly want to
accept NIBI into the fold—an attitude which at present is not evident. Also, if NIBI is to succeed
there must be advance treaties with the organ/disease institutes.
The academic radiology community‘s enthusiasm for supporting and conducting strong basic
research has been growing steadily over the past 20 years. Thus it is natural to think about having
one‘s own institute. However, in reality, there may be more opportunities within well-funded
existing programs than in a startup institute. The risk is in creating an institute in name alone
without the prerequisite resources to enable an effective increase in biomedical imaging research.
Perhaps the more logical approach would be to look harder for those opportunities. For example,
NCI‘s relatively new Diagnostic Imaging Program presents untapped potential. The startup time
for a new NIH entity will be prolonged. One might expend the same effort and achieve greater
results within the existing structure.
I, for one, do not feel that the present funding of medical imaging research should be described
as ‗‗chaos‘‘ induced by lack of an effective voice at the top. Efforts to better conduct the
assembled chorus may serve to produce a harmony of results far in excess of those envisioned
through NIBI. The NIH should represent science and clinical innovation and not be defined
within narrow technologies. Medical physicists might well benefit most through reorchestrating
communications within the NIH rather than through formation of a new institute.
443
10.9. Candidacy for Board certification in Radiological/Medical

Physics should be restricted to graduates of accredited training
programs
Timothy K. Johnson and Bhudatt R. Paliwal

Reproduced from Medical Physics, Vol. 27, Issue 5, pp. 825-827, May, 2000
OVERVIEW
In medicine, physician candidates for certification by boards recognized by the American Board
of Medical Specialties must be graduates of training programs accredited by the Accreditation
Council for Graduate Medical Education. There are two certification boards for medical
physicists, the American Board of Radiology and the American Board of Medical Physics.
Neither board requires graduation from an accredited graduate or residency program for
individuals admitted to candidacy for board certification. As a consequence, there is little
incentive for program directors or their institutions to seek accreditation of graduate or residency
programs from the Commission on Accreditation of Medical Physics Education Programs
(CAMPEP). Also, the dissociation between accreditation and certification in medical physics
ignores one of the major criteria for assuring the quality of practitioners in a healthcare
discipline. However restricting certification eligibility to graduates of accredited programs
would penalize individuals who continue to enter medical physics in unconventional ways that
would never be considered appropriate for accreditation. How this dilemma should ultimately be
resolved is the subject of this Point/Counterpoint.
Arguing for the Proposition is Timothy K. Johnson, Ph.D. Dr. Johnson is Associate Professor in
the Department of Radiology, and Director of the Graduate Program in Medical Physics at the
University of Colorado Health Sciences Center. He is the author of the MABDOSE internal
radionuclide dosimetry software [see Med. Phys. 26, 1389–1403 (1999)], and has high hopes for
radioimmunotherapy as an alternative to external beam and chemotherapy in the treatment of
cancer.
Arguing against the Proposition is Bhudatt R. Paliwal, Ph.D. Dr. Paliwal is a Professor of Human
Oncology and Medical Physics at the University of Wisconsin, Madison. He joined the
University of Wisconsin in 1973 and has served as the Director of Radiation Oncology Physics
for more than 20 years. Dr. Paliwal has served on many AAPM committees, including AAPM
committees, including the office of president in 1996. Currently, he is the chairman of the
Commission on Accreditation of Medical Physics Education Programs (CAMPEP, Inc.) and the
American College of Medical Physics. He is also a trustee of the American Board of Radiology.
Dr. Paliwal is certified by the ABR and ABMP.
FOR THE PROPOSITION: Timothy K. Johnson, Ph.D.
Opening Statement
444
When a profession establishes itself as an entity, it does so with the assumption that its qualities,
characteristics, goals, and aspirations are unique. According to Webster‘s definition of
profession, assumption of these characteristics entails advanced education and training. The
natural progression of a profession, as more and more individuals enter it, is for the founding
members to create some method to ensure that new members really possess the traits that define
the profession. The difficulty encountered by various AAPM committees, when trying to define
what a ‗‗trained medical physicist‘‘ is, simply emphasizes the emotional nature of this issue.
In medicine, not just anyone can sit for the United States Medical Licensure Exam. You must
have graduated from an accredited medical school. Similarly, you cannot practice law without
passing a bar exam. With the exception of California, you cannot sit for the bar without having
graduated from an accredited law school. California administers a pretest to assess competency
of non-graduates to sit for the bar exam. While a pretest is conceivable for Medical Physics
certification, it would require more time and effort from the testing bodies than they are probably
willing to commit.
The primary reason for restricting board certification to graduates of accredited medical physics
programs is efficiency. The time and effort required of volunteers who implement and proctor
the examinations would be minimized. This is because graduation from an accredited program
implies that a core body of knowledge has been received. Mastery of this knowledge can be
tested in a cursory fashion through written and oral exams for certification. It cannot be
exhaustively tested, because the body of knowledge is extensive. Additionally, the breadth and
scope of problem-solving ability is not encapsulated in a written exam or thoroughly probed in
an oral exam; again, the knowledge required to practice medical physics is too great. With
certification linked to accreditation, the certification body effectively becomes a partner with the
accrediting agency in promoting individuals to practice professionally as medical physicists.
It is certainly possible to have an open certification process so that graduates of unaccredited

programs can become certified. Unfortunately, this opens a Pandora‘s Box for a curriculum to be
called a ‗‗Medical Physics‘‘ education (e.g., a graduate physics program that includes a single
image reconstruction course, or a terminal master‘s program for medical dosimetrists). To
rigorously test graduates of such programs so that a minimum level of competence is guaranteed
would require an inordinate amount of time. Most professions have addressed the certification
problem by requiring graduation from an accredited program as a minimum standard for entering
the profession. This requirement does not imply that an individual with a different educational
background could not pass the exam. It simply provides a reasonable degree of assurance that
with multiple checkpoints and reasonable time expenditures incompetent individuals are not
promoted to jobs where they could injure people. This two-tier approach provides a level of
redundancy in examining individuals and weeding out incompetents. In so doing, it protects
patients and the integrity of the profession.
Rebuttal
Dr. Paliwal and I agree that the primary objective of certification is to assess minimum
competence. Dr. Paliwal however, asserts that certification is an acknowledgment that the
governing board has examined the candidate thoroughly in all knowledge appropriate to the
profession. This is unlikely given the constraints in the exam format currently administered by
the American Board of Radiology. Although the scope of the physics exams (written and oral)
offered by the American Board of Radiology is broad, I recall at the conclusion of taking both
exams the number of topics that were NOT covered (some with a sigh of relief).
445
The two issues that Dr. Paliwal presents are in fact hypothetical. Regarding the issue of
candidates from foreign countries, the educational background of foreign applicants could be
reviewed on a country-by-country basis for program similarities to accredited U.S. programs.
Akin to physician licensure, individuals graduating from programs whose content is substantially
similar to that embodied in AAPM Report No. 441 could be allowed to sit for board certification.
Regarding the issue of supply and demand of medical physicists, one doesn‘t address the
problem by opening the floodgates to any and all comers. Rather than reducing standards, efforts
should be expended in supporting and possibly expanding accredited programs to accommodate
additional students. Instead of creating a candidate shortage, restricting candidacy would
increase the census of students enrolled in accredited programs.
Relying on an accredited program as a requirement for licensure provides additional screening of

a candidate‘s background. Since an ill-prepared individual could do irreparable harm to patients,
extra caution should be used to guard against the unintentional promotion of incompetence.
Requiring individuals who sit for board certification to come from an accredited program or its
foreign equivalent is the most reasonable way to ensure this.
AGAINST THE PROPOSTION: Bhudatt R. Paliwal, Ph.D.
Opening Statement
The primary objective of certification is to assess the minimum competence of candidates

involved in the practice of a profession. The issuance of certification by an examining Board is
an acknowledgment by the Board that it has examined the candidate thoroughly in all aspects of
basic and applied knowledge important to the profession. The Board must be fair in its practice
and provide equal opportunity to all individuals who seek to establish their level of competence.
The restriction of candidacy for board certification to graduates of accredited training programs
would not provide a fair opportunity to individuals who did not graduate from an accredited
program. Hence, limiting certification to candidates from accredited programs would be
discriminatory to many highly competent and qualified physicists who are trained in
nonacademic clinical environments.
This approach would also have an extremely adverse effect on our profession. There are only 11
accredited training programs, and one of these is in Canada. On the average, these programs
produce about 120 medical physicists per year. This is about 2.5% of the current AAPM
membership. Some of the new graduates are highly specialized in subspecialties of medical
physics. Hence, limiting candidacy to accredited program graduates would create a shortage of
certified candidates.
A large number of medical physicists, who over the years have made a significant contribution to
the development of our profession, have come from many foreign countries and from
nontraditional academic programs. The institutions where they were trained were not accredited
and are not likely to be accredited in the future. They are unlikely to be able to participate in the
accreditation process because of a lack of financial resources or other practical considerations.
Consequently, limiting candidacy to accredited programs would be discriminatory to foreign,
well-qualified individuals.
446
Above all, I believe a competence-based open certification process, allowing everyone to

establish his or her credentials, is fair. It enhances competition, enriches the field, and contributes
to high-quality patient care.
Rebuttal
It is true that Webster‘s definition of a profession and the creation of founding fathers are the
guidelines we use to keep ourselves on a narrow and focused path. However, the changes and
suggestions made by AAPM committees are usually not emotional in nature. They have
relevance to the evolutionary nature of our profession and reflect the adjustments we need to
make to keep our profession flourishing and our professional standards in synchrony with the
changes taking place around us.
A candidate who has graduated from an accredited program does carry a seal of achievement of a
minimum standard of education. In spite of the accreditation process, we find a significant
diversity in the breadth and scope of education and training provided by accredited programs.
Accreditation has not reached the level of standardization and specificity found in the legal and
medical professions. The profession of medical physics is not bound by any geographic
boundaries. We cannot even begin to compare ourselves to the legal and medical professions in
terms of numbers. They deal with thousands of candidates each year, whereas we examine less
than one or two hundred per year. If we were to plot a histogram of numbers of programs for
each profession (medical, legal, and medical physics), medical physics would be at the same
level as the x axis. We are not exactly being swamped by the number of candidates taking the
board exams nor have we a surplus of medical physicists.
Our profession is undergoing substantial change in its scope and in many of its subspecialties,
particularly in medical imaging, applications of lasers, and integration of computerized systems.
AAPM has just formed an ad hoc committee to assist in the formation of a professional group for
radiotherapy accelerator engineers, hopefully within the confines of AAPM. In order to attract
new talent, we need to keep the pathways open.
An open certification allows the graduates from well-established programs in the physical
sciences to prepare themselves for certification. It also provides candidates from North America
and other developed countries an opportunity to establish their credentials. If they meet the
specified standards they should be welcomed to the profession to share their expertise, help the
profession provide the necessary resources to the community and contribute to the growth of the
profession.
REFERENCE
1. AAPM Report No. 44, ‗‗Academic program for master of science degree in medical
physics‘‘ (American Institute of Physics, Woodbury, 1993).
447
10.10. Graduation from a CAMPEP or equivalent accredited

graduate or residency program should be a prerequisite for
certification in radiological physics
Michael G. Herman and Howard Amols

OVERVIEW
Virtually all American Board of Medical Specialties (ABMS)-associated certification boards

require graduation from an accredited training program as an eligibility requirement for a
certification examination. One exception is radiological physics, where eligibility is based
principally on a graduate degree in physics or related science or engineering discipline, plus three
years of clinical experience. Whether graduation from an accredited graduate or residency
program in medical physics should be a prerequisite for certification in radiological physics is the
topic of this Point/Counterpoint.
Arguing for the Proposition is Michael Herman, Ph.D. Dr. Herman earned a Ph.D. in
experimental nuclear physics in 1986. In 1989 he joined the Radiation Oncology staff at Johns
Hopkins. He served on the faculty and in the capacities of Acting Chief and Associate Director of
Medical Physics. In 1998, Dr. Herman joined the Division of Radiation Oncology at Mayo Clinic
Rochester. He is currently Head of Physics, Associate Professor and full member of the graduate
faculty. He is ABMP certified with an ABR letter of equivalence. Dr. Herman serves actively in
the AAPM, ACMP, CAMPEP and ABR. He is a fellow of the AAPM and ACMP.
Arguing against the Proposition is Howard Amols, Ph.D. After receiving a non-CAMPEP
approved Ph.D. in Nuclear Physics from Brown University in 1974, Dr. Amols did a non-
CAMPEP approved post-doc at Los Alamos National Laboratory. He has held medical physics
positions at the University of New Mexico, Brown, and Columbia Universities where he also
taught numerous non-CAMPEP approved courses in medical physics. He is currently Chief of
Clinical Physics at Memorial Sloan Kettering Cancer Center which, perhaps not surprisingly, has
a non-CAMPEP approved post doctoral training program. He is certified by the ABMP, and
along with 265 other non-CAMPEP graduates is a Fellow of the AAPM. On January 1, 2005 he
became the 46th consecutive non-CAMPEP approved President of the AAPM.
FOR THE PROPOSITION: Michael Herman, Ph.D.
Opening Statement
The Institute of Medicine publication "To Err is Human" substantially raised the profile of the
costs of medical errors.1 The profession of medical physics is not exempt from the necessary
accountability this report and the public demand. Yet, due to many conditions, too many
individuals are placed in (independent) medical physics services without the necessary training
and experience to discharge clinical duties safely and competently. This practice results in
448
unacceptably low board scores, gives medical physics a bad name and worst of all, can greatly
compromise patient care.
Quality patient care is paramount, and thorough clinical training is essential for individuals
entering practice of clinical medical physics. The common premise among all 24 certification
boards (including the American Board of Radiology (ABR)) of the American Board of Medical
Specialties is "To achieve initial certification, each board requires between 3 and 6 years of
training in an accredited training program and a passing score on a rigorous cognitive
examination."2 Former ABR president Casarella agrees that the certification exam alone cannot
cover every area of practice and that all candidates sitting for the exam must have received
proper training in the essentials of practice in an accredited program. "It is the successful
completion of the residency itself that is the sine qua non (absolute prerequisite) of ABR
certification."3 Why should medical physics be different? The knowledge required to practice
medical physics is too extensive to be completely evaluated in the certification process.4
Structured training like a residency, followed by a certification exam represents the pathway by
which a clinical medical physicist can be expected to develop the breadth and depth of
knowledge to independently discharge clinical duties.
CAMPEP accredited medical physics residencies provide the absolute prerequisite training that,
when followed by board certification, produce individuals properly prepared to practice medical
physics. Unfortunately, there is a major shortage of qualified medical physicists and a dearth of
accredited training programs to produce these individuals. These topics have been discussed
fervently in recent AAPM newsletters. Raising the bar on the ABR exams could reduce the
number of improperly trained individuals entering medical physics, but would exacerbate the
physicist shortage.5
To maintain quality and increase supply, additional opportunities for structured training must be
developed. These can take the form of a residency, a graduate or an on the job training program,
as long as the essential prerequisites (as detailed in AAPM report #36)6 are learned and
CAMPEP provides accreditation. This broader model increases available programs (and
graduates) and maintains the intellectual diversity that has historically been a strength of medical
physics. AAPM and other organizations must support an increased focus on structured training
and CAMPEP must facilitate accreditation of these programs. Certification bodies should accept
only graduates of these programs (the ABR physics trustees suggest by 2012). The quality of
patient care, our accountability to the public, and our stature among medical professionals are at
stake.
Rebuttal
My Point/Counterpoint colleague and I agree that proper training for medical physicists is
essential and that board certification combined with proper training instills confidence that
individuals will enter the practice of clinical medical physics in a competent manner.
As noted, however, certification alone is not the answer, because it is impossible to test the
breadth and depth of one's training in such an exam, and raising the bar serves only to reduce the
number of qualified individuals but does little to improve the quality of candidates.
Dr. Amols requested statistics that suggest that CAMPEP-trained individuals fare better on the
board exams. I am pleased to oblige. Nine CAMPEP-accredited residency programs have
produced 54 graduates, of which 38 have taken the full certification exam. Of these, 36
449
candidates passed the full board exam on the first attempt. Thus the first time pass rate for
graduates of CAMPEP-accredited residency programs is 95%. 7 The overall ABR physics pass
rate in recent years is 53%, including first-through third-time takers.8 I also examined for the
ABMP for many years and the average pass rates for ABMP are similar to if not lower than the
ABR values. I conclude from these data that a 95% pass rate for the subset of candidates trained
in CAMPEP programs is significantly better and different than the average certification exam
pass rates of ~50% for all applicants.
It is essential for AAPM and other organizations that represent clinical medical physicists to
make it policy to support and facilitate CAMPEP-accredited structured training as THE
requirement to sit for board exams. Completion of this pathway qualifies individuals to discharge
duties in the clinical environment. Any program that meets the requirements for clinical training
as outlined by CAMPEP and AAPM report number 36 should be accredited. Better and
consistent training brings improved performance that is good for patient care and good for
medical physics.
AGAINST THE PROPOSITION: Howard Amols, Ph.D.
Opening Statement
CAMPEP approved programs are good. We debate only whether it should become the only
pathway into medical physics. For 100 years we've had multiple pathways; a degree in physical
science or engineering, on the job training, fellowships, and more recently residencies. Whatever
path taken, a peer-reviewed certification exam becomes the equalizer. This system has flaws but
basically works and has brought intellectual diversity to our profession, which is one if its
greatest strengths. "If it ain't broke don't fix it."
Yes, we have a manpower shortage, and many unqualified candidates pass the board exams. But
restricting entry to the boards (and therefore the profession) only to CAMPEP graduates is not
the best way to solve either problem. We can more fairly reduce the number of unqualified
people passing the boards simply by failing everyone who deserves to fail. Make the questions
harder and/or raise the passing score. It's that easy! CAMPEP can't fix the boards; training and
certification are separate issues! In my 10 years experience as an ABMP examiner I've seen no
evidence that a candidate's performance is correlated with CAMPEP training. I challenge my
opponent to produce statistics to the contrary.
What about the manpower shortage? Restricting entry into medical physics only to CAMPEP
graduates will only make this problem worse. There simply are not enough CAMPEP programs
to meet manpower needs. I cannot accept the Cinderella fantasy we hear that by making
CAMPEP mandatory the needed programs will magically materialize. There isn't enough money
and there aren't enough teachers.9,10 Even if there were, a CAMPEP only policy would divest our
profession of intellectual diversity and creativity.
Another argument we hear in favor of CAMPEP is that other ABMS boards have similar
requirements. So what? Other ABMS boards require students to dissect cadavers. Should we
require that also? We are physicists, and no other scientific specialty has such a restrictive
requirement. What about the argument that medical physics has become soooooo complicated
that CAMPEP training is the ONLY way to master it? This argument is historically unjustifiable,
self-serving, and insulting to thousands of competent medical physicists who've succeeded
450
without CAMPEP. Even if one believes this myth, what harm is done by letting non-CAMPEP
people take the boards? If they're really incompetent they'll fail. Further, the CAMPEP only
proposal makes no distinction between graduate programs and residencies. According to their
own guidelines:11 "CAMPEP accreditation of a program does not address the clinical competency
of individual graduates. Certification that an individual medical physicist has demonstrated a
prescribed level of professional competence is currently available from the American Board of
Radiology."
In a recent survey of 2100 AAPM members12 91% supported board certification and 86%
supported CAMPEP programs in principle, but only 21% supported barring non-CAMPEP
graduates from the process, and only 14% supported CAMPEP residency as the only entryway.
Support CAMPEP, yes. Only route possible, NO!
Rebuttal
Medical physics faces two major problems: a manpower shortage, and physicists with incomplete
clinical training passing the boards. Neither problem has anything to do with CAMPEP. The
former results from poor recruiting, and the latter from a board exam that is arguably not
comprehensive enough in scope. Dr. Herman confuses these issues and tries to fix both by
making CAMPEP mandatory. Historically, people have achieved competency and passed the
boards without CAMPEP. Making CAMPEP mandatory exacerbates the manpower shortage,
doesn't address the comprehensiveness or rigor of the boards, and without large infusions of
money does not address the `dearth of accredited training programs' that he refers to either.
Worse, it effectively bars entry into our field to a variety of bright people with broad ranging
skills.
I agree with Dr. Herman that clinical training is essential and improper training compromises
patient care, but the rigid framework of CAMPEP is not the right answer. Even CAMPEP's own
guidelines state that "demonstrat[ion of] a prescribed level of professional competence is
currently available from the American Board of Radiology (ABR)"11 not CAMPEP.
Indeed, if this were not true there'd be no need for CAMPEP graduates to take the board exam! In
his opening statement Dr. Herman said nothing about CAMPEP degree programs, I think because
he and I agree that a CAMPEP degree guarantees little about clinical competency. He therefore
limits his remarks to CAMPEP residency programs which do focus on clinical training. But he is
again stymied by the "dearth of accredited training programs."
In summary, CAMPEP degrees offer little or no added value over other physics or engineering
degrees with regard to clinical competency, and CAMPEP residencies cannot meet the
manpower needs. For these reasons, as currently formulated, making CAMPEP mandatory for
entry into medical physics is a recipe for disaster.
In his closing paragraph Dr. Herman states that "additional opportunities for structured training
must be developed." I agree. I think a better solution is for AAPM, CAMPEP, and ABR to
defocus on degree programs and instead recognize alternate pathways to formal residencies. For
example, CAMPEP approved on the job training under the guidance of accredited mentors
similar to the old guild system where one apprentices under a master, becomes a journeyman,
and ultimately a master. If accreditation procedures can be agreed upon along these lines as
alternatives to formal institutionalized CAMPEP programs I would be in favor.
451
REFERENCES
1. K.T. Kohn, J.M. Corrigan, and M.S. Donaldson, To Err is Human: Building a Safer
Health System (Institute of Medicine, 1999).
2. T. A. Brennan, R. I. Horwitz, F. D. Duffy, C. K. Cassel, L. D. Goode, and R. S. Lipner,
"The role of physician specialty board certification status in the quality movement," JAMA, J.
Am. Med. Assoc. 292, 1038–1043 (2004).
3. W. J. Casarella, "Current structure and purpose of the American Board of Radiology
examination process," Acad. Radiol. 8, 1260–1261 (2001).
4. T. K. Johnson and B. R. Paliwal, "Candidacy for board certification in
radiological/medical physics should be restricted to graduates of accredited training programs,"
Med. Phys. 27, 825–827 (2000).
5. W.R. Hendee, "Linking Accreditation and Certification in Medical Physics," Journal of
the American College of Radiology 2, 198–199 (2005).
6. E. Sternick, R. Evans, E. Heitzman, J. Kereiakes, E. McCulough, R. Morin, T. Payne, J.
Purdy, and N. Suntharalingam, "Essentials and Guidelines for Hospital Based Medical Physics
Residency Training Programs," AAPM Report #36 (1990).
7. Personal Communication, "Survey of CAMPEP accredited radiation oncology program
directors," (2004).
8. W.R. Hendee, "Accreditation, Certification and Maintenance of Certification in Medical
Physics: The Need for Convergence," NCCAAPM meeting, Nov. 19 (2004).
9. AAPM Newsletter 29, No. 1, pp. 22–23 (2004).
11. Guidelines for Accreditation of Graduate Educational Programs in Medical Physics,
Commission on Accreditation of Medical Physics Educational Programs, www.campep.org
(1998).
452
10.11. Deployment of a Maintenance of Certification program in

medical physics serves the interest of the profession and the public
Stephen R. Thomas and Joseph S. Blinick

OVERVIEW
Maintenance of Certification (MOC) is being implemented in most medical specialties as a

follow-up to time-limited certifications. In the radiological disciplines, the MOC process is
administered by the American Board of Radiology, and focuses on lifelong learning, continuous
self-assessment, and periodic examination. Some medical physicists believe this process will
upgrade the profession by encouraging physicists to remain current with their discipline. Others
feel that staying up-to-date in the field is a professional and ethical obligation that practicing
physicists are already pursuing, and that oversight by a professional organization is not required.
This difference of opinion is the topic of this month's Point/Counterpoint.
Arguing for the Proposition is Stephen R. Thomas, Ph.D. Dr. Thomas is Professor of Radiology
at the University of Cincinnati College of Medicine and Director of Medical Physics for
University Radiology Associates. He serves as an ABR Trustee in Radiologic Physics (Medical
Nuclear Physics) and is Chair of the ABR MOC Coordinating Committee. He is an elected
member of the ABMS Committee on Oversight and Monitoring of MOC. Dr. Thomas has been
active within the RSNA and currently is a trustee of the RSNA Research and Education
Foundation. He is a Fellow of the AAPM and has participated at many levels within that
organization, including President in 1997.
Arguing against the Proposition is Joseph S. Blinick. Dr. Blinick received his Ph.D. in Physics
from Brown University in 1971 and began his career in medical physics at Boston City and
University Hospitals in Boston. From 1974 until his retirement in 2001, he was Chief Radiation
Physicist and RSO at the Maine Medical Center in Portland. Since that time he has continued to
work as a consultant. He is a former Treasurer of the AAPM and also served as Placement
Service Director. Dr. Blinick was certified by the ABR in Radiological Physics in 1976. He was
also certified by the ABMP in Radiation Oncology Physics in 1991 and was recertified in 2001.
He served for two years as an oral examiner for the ABMP in Medical Health Physics. He is a
Fellow of the AAPM, the ACMP and the ACR.
FOR THE PROPOSITION: Stephen R. Thomas, Ph.D.
Opening Statement
In a major paradigm shift for medicine, the previously established mode of lifetime certification
for healthcare professionals based upon a one-time successful passing of a cognitive examination
has been replaced by time-limited certification incorporating a program of continuous
professional development. This new approach has been embraced by all 24 member boards of
The American Board of Medical Specialties (ABMS), with the boards being committed to
453
implementation of an appropriately designed Maintenance of Certification (MOC) program in

their specialties and sub-specialties. Across the health care industry there has been a growing
awareness that continuous professional enhancement is essential to ensuring high quality health
care in this era of rapidly advancing technology.1,2 The shift toward MOC is in response to the
recognition that active programs and evidenced-based measures are required to demonstrate that
practitioners are remaining current in their fields. In brief (details may be found in Ref. 3), the
MOC program entails requirements related to the 4 components (Professional Standing; Life-
long Learning and Self-Assessment; Cognitive Expertise; Assessment of Performance in
Practice) linked to the 6 competencies (Medical Knowledge; Patient Care; Interpersonal and
Communication Skills; Professionalism; Practice-Based Learning and Improvement; Systems-
Based Practice) expected of healthcare professionals.
I am certain medical physicists would agree with the proposition that remaining up-to-date in
one's discipline should be mandatory, both to satisfy the public trust and as a matter of
professional responsibility. It is the means for accomplishing this goal that is the topic of
discussion here. Public and professional confidence can be established and most efficiently
maintained through unified programs that are sanctioned by the vested organizations. For
radiology, that organization is the American Board of Radiology (ABR) which, as the
certification Board for radiological physicists, has the vision to ensure that its diplomates possess
the requisite knowledge, skills, and experience to provide high-quality patient care. Importantly,
the MOC components and competencies were not formulated independently, but represent the
consensus of all 24 medical boards as coordinated and approved by the ABMS. Thus all
healthcare constituents can be assured that the MOC programs were carefully developed, unified
as to concepts, and extend with consistency across all disciplines of medicine. This mechanism
obviates any concerns that standards might not have been appropriately defined and executed for
any specific individual. In a very real sense, coordination by the ABR assists the medical
physicist through presentation of an unambiguous pathway for progression toward achieving the
goal of documenting continuous professional enhancement. In the real world, characteristic
vagrancies appear when a myriad of individually prescribed programs evolve independently.
Without this integrating structure, the process runs the risk of being considered meaningless, and
might ultimately come under the threat of external regulation.
As a positive logistical aspect promoting efficiency of the MOC process, the ABR will be
expanding its website to provide personal, secure files for diplomates to assemble data and
documents over the 10-year cycle. This will serve as a repository for CME credits, professional
standing citations, cognitive exam results, etc. If desired by the diplomate, the ABR will assist in
review of the diplomate's "MOC portfolio" at various stages in the cycle and offer helpful advice
as progress is evaluated.
In summary, medical physicists will benefit through participation in a unified MOC program
coordinated by the ABR that conforms to a national standard.
Rebuttal
In responding to my colleague, let me first address the IOM report on medical errors.4 This report
is focused exclusively on patient safety. The MOC process incorporates segments associated with
this important topic. However, MOC has the broader mission to ensure progression toward
higher-quality health care through documentation of diplomate participation in programs relevant
to the evolution of technology in health care. Although the term "MOC" may not appear
explicitly within the IOM report, many of the recommendations would be addressed through
454
implementation of MOC programs. For example, Recommendation 7.2 on performance standards

for health professionals calls on licensing bodies to "Implement periodic re-examinations and re-
licensing of doctors, and other key providers." Clearly, MOC is positioned to play a crucial
role in achieving the intent of this recommendation.
The issue of requiring MOC for diplomates holding lifetime certificates does deserve
consideration. Because of legal constraints, the ABR cannot change the original terms of
issuance. The ABR is actively encouraging all of its "older' diplomates to voluntarily enter MOC.
All Trustees of the ABR have enrolled. Through a personal sense of professional responsibility
and potential requirements of employers that medical physicists formally engage in MOC, it is
the expectation that a majority of diplomates in active practice will sign up.
Dr. Blinick acknowledges the significance of assessing performance in practice (4th component
of MOC). At the end of Dec. 2004, all Boards submitted Part 4 plans that currently are under
evaluation by the ABMS. The challenge of measuring efficacy in these programs is recognized,
but the expertise required is available. There are never absolute guarantees of outcome, but
providing tools for assisting performance assessment represents a positive step forward.
The statistics quoted on the number of full AAPM members who do not hold certification require
analysis. Medical physicists with an advanced degree working directly in a clinical environment
undoubtedly are certified (or are planning to be). These are the individuals for whom MOC is
intended. Other valued members of the AAPM include individuals with undergraduate degrees
only, those waiting to qualify for certification, professionals not directly in patient care (basic
scientists) and foreign nationals. There are valid reasons why certification may not be applicable
under specific situations.
It is not claimed that an ABR-MOC certificate is a guarantee of competence. Rather, it

documents that the diplomate has satisfied the requirements of the MOC program (including
mastery of a body of knowledge and skills with accountability in their application to patient
care), which can be considered an indication of the commitment to maintaining competency. This
is a good thing!
AGAINST THE PROPOSITION: Joseph S. Blinick, Ph.D.
Opening Statement
A recent journal article3 described the American Board of Radiology (ABR) Maintenance of
Certification (MOC) program, which is designed to respond to public demand that the safety and
quality of American medicine be improved.
The document states that Board certification has been accepted as a good, albeit imperfect,
process. But now "questions have been raised as to whether professionals with lifetime
certificates maintain the knowledge, skill and familiarity necessary to continue providing quality
patient care." In fact, there is no evidence that certification provides any assurance of quality,
much less that the MOC program is really necessary at this time. A report of the Institute of
Medicine (IOM) estimated that between 44,000 and 98,000 people die each year in American
hospitals due to medical errors.4 Most of these errors are undoubtedly committed by licensed
and/or certified practitioners. Furthermore, during the period in which the data were collected
several member boards of the American Board of Medical Specialties (ABMS) had already
455
instituted recertification programs for their diplomats,5and evidence of continuing education had
been required by most certification boards and state licensing boards for many years. The IOM
had many suggestions for reducing medical errors including establishment of a Center for Patient
Safety, a nationwide mandatory reporting system of medical errors with protection from lawsuits,
and a commitment by professional societies to establish committees on patient safety. Nowhere is
there a recommendation for maintenance of certification.
Even if we assume that MOC is desirable, the current program will not be effective. For one
thing, it is voluntary for those certified before 2002. If the ABR really thinks older practitioners
aren't keeping up, the MOC program should be required for all certificate holders regardless of
when they were certified. In addition, MOC still relies heavily on amassing huge, but arbitrary
numbers of continuing education units (CEU) and taking multiple cognitive examinations over a
ten year period. Rockhill6 states that "mandatory continuing education may actually limit
learning." She advocates leaving control of continuing education with the individual. Miller 7
points out that knowledge of a field (which education provides and cognitive tests assess) is
necessary but not sufficient. Professionals have to know how to apply this knowledge.
To its credit, the ABR has included a section on assessing performance in practice. Regrettably, it
has left this entirely undefined for the present. Performing this type of assessment effectively will
be both very difficult and very expensive. There have been many methods developed to measure
performance in the past,8 but none has proven to be universally reliable. Unfortunately, even if a
professional can show that he/she knows how to apply knowledge, it doesn't guarantee that
he/she will always do so properly in real situations.
Approximately 90% of radiologists are certified by the ABR, 9 but according to information
provided by Michael Woodward of the AAPM, only 56% of AAPM full members are certified
by the ABR or ABMP. Are the remaining 44% therefore incompetent? If one believes that
certification is important to our profession, the priority ought to be to convince more physicists to
become certified. In this regard, the complexity, cost and burdensome record-keeping of the
MOC program are counter-productive.
The MOC program will not result in more competent physicists, nor will it promote
professionalism among physicists.
It does, however, provide the perception that we are "doing something" about the problem of
incompetent professionals.
Rebuttal
Since several medical boards have had time-limited certification for some time, the MOC
program is not so much a paradigm shift as a burdensome and expensive addition to existing
programs. Rather than actually addressing the question of professional incompetence, it seems
primarily aimed at "achieving the goal of documenting continuous professional enhancement,"
and preventing "the threat of external regulation."
The program still relies heavily on accumulating mandatory CEUs and taking cognitive
examinations, neither of which assures continued professional competence. Further, it is one
thing for a board to require mastery of a well-defined body of knowledge for initial certification,
but it's quite another for it to define what each practitioner currently needs to know after several
years of post-certification career divergence have gone by.
456
The "Assessment of Performance in Practice" component is still completely undefined. It could

be important if done right. But to do it right will be very difficult and expensive.
The remaining component, "Professional Standing" requires, among other things, documentation
of "expertise-based appointments" to groups such as the NCRP, ICRP and NIH study sections or
hospital medical staffs. This clearly favors academically based physicists who have both the time
and the financial backing to participate. But it's hard to see how these activities actually assure
clinical professional competence.
The MOC program fails to actually address the issue of continued professional competence and
its cost and complexity make it an unnecessary burden for the relatively small number of medical
physicists who are currently certified.
REFERENCES

"The role of physician specialty board certification status in the quality movement," J. Am. Med.
Assoc. 292, 1038–1043 (2004).
2. J. E. Madewell, "Lifelong learning and the maintenance of certification," J. Am. Coll.
Radiol. 1, 199–203 (2004).
3. S. R. Thomas, W. R. Hendee, and B. R. Paliwal, "The American Board of Radiology
Maintenance of Certification (MOC) Program in Radiologic Physics," Special Topic, Med. Phys.
32, 263–267 (2005).
4. To Err is Human—Building a Safer Health System, edited by L. T. Kohn, J. M. Corrigan,
and M. S. Donaldson (National Academy Press, Washington, D.C., 2000).
5. 2004 Annual Report and Reference Handbook (American Board of Medical Specialties,
Evanston, IL).
6. K. Rockhill, "Professional education should not be Mandatory," in Examining
Controversies in Adult Education, edited by B. W. Kreitlow and Associates (Jossey-Bass
Publishers, San Francisco, CA, 1981).
7. G. E. Miller, "The assessment of clinical skills/competence/performance," Acad. Med.
65, S63–S67 (1990).
8. D. B. Swanson, G. R. Norman, and R. L. Linn, "Performance-based assessment: lessons
from the health professions," Educational Researcher 24, 5–11 (1995).
9. J. H. Sunshine, Y. S. Cypel, and B. Schepps, "Diagnostic radiologists in 2000: Basic
characteristics, practices, and issues related to the radiologist shortage," AJR, Am. J.
Roentgenol., 178, 291–301 (2002).
457
10.12. The proposed MOC requirements for physicists are

reasonable
Stephen R. Thomas and Jeffrey A. Garrett
OVERVIEW
It is important for specialists in the medical professions to ensure that new technological
developments are continually being incorporated into their practice and made available to
patients. This is especially important for medical physicists because they are in the forefront of
development and introduction of these new methodologies. Consequently the American Board of
Radiology (ABR) has initiated a new Maintenance of Certification (MOC) process that is
mandatory for all those medical physics diplomates holding time-limited certificates, and highly
recommended for all others. The requirements of the MOC program include, over the 10-year
MOC cycle, earning 500 continuing education credits, completion of 20 self-assessment modules
(online multiple-choice "tests"), and successful completion of a closed-book, proctored multiple
choice exam in the 8th, 9th, or 10th year of the MOC cycle.1 The proposed annual fee for
maintenance of certification is $170. These are extensive requirements that some believe are
excessive and/or unfair. Others argue, however, that they are essential if we are to assure that all
diplomates keep up-to-date in their knowledge so as to promote the best interests of patients.
This disagreement is the topic of this month's Point/Counterpoint debate.
Arguing for the Proposition is Stephen R. Thomas, Ph.D. Dr. Thomas is Professor of Radiology
at the University of Cincinnati College of Medicine and Director of Medical Physics for
University Radiology Associates. Within the ABR, he has recently assumed the position of
Associate Executive Director (Radiologic Physics). Through June 2006, he will continue to serve
as an ABR Trustee in Radiologic Physics (Medical Nuclear Physics) and Chair of the ABR MOC
Coordinating Committee. He is an elected member of the American Board of Medical Specialties
(ABMS) Committee on Oversight and Monitoring of MOC. Dr. Thomas has been active within
the RSNA and currently is a Trustee on the RSNA Research and Education Foundation. He is a
Fellow of the AAPM and has participated at many levels within that organization including
President in 1997.
Arguing against the Proposition is Jeffrey A. Garrett, M.S. Mr. Garrett is the Chief Medical
Physicist and Radiation Safety Officer for Mississippi Baptist Medical Center in Jackson, MS.
Mr. Garrett earned a B.S. degree in Physics at The Citadel and an M.S. in Medical Physics from
East Carolina University in 1997. Upon graduating from East Carolina, Mr. Garrett worked as a
Junior Physicist in Albuquerque, NM. From there he moved to Jackson, MS and was employed
as a Staff Physicist until eventually assuming the role of Chief Physicist in 2001. Mr. Garrett was
certified by the ABR in 2002 and currently serves on the AAPM MOC Task Group No. 127.
FOR THE PROPOSITION: Stephen Thomas, Ph.D.
Opening Statement
458
In March 2000, all 24-member boards of the American Board of Medical Specialties (ABMS)
agreed to implement MOC. The mandate is that each board institute programs that include four
components: professional standing; lifelong learning and self-assessment; cognitive expertise;
and assessment of performance in practice. 2,3,4 Although individual boards have some flexibility
in designing methods to comply, programs undergo rigorous scrutiny before receiving ABMS
approval. The ABMS Committee on Oversight and Monitoring of MOC will hold boards
accountable for adherence to their programs.
The ABR did not develop MOC requirements in a vacuum. Input and guidance were solicited
from all sponsoring societies at a series of meetings hosted by the ABR (December 2001(Tucson,
AZ): ABR Meeting on Maintenance of Certification; January 2004 (Tucson, AZ): Implementing
MOC: Issues and Strategies; August 2005 (Chicago, IL): Self-Assessment Modules—Summit
Meeting). It was recognized that the program elements must be constructed such that
requirements could be satisfied in a reasonable fashion. Working groups including the AAPM
Task Group on MOC and the RSNA MOC Coordinating Committee will play important roles in
helping to shape programs as the process evolves.
Professional standing may be satisfied in a straightforward fashion through documentation of a

valid, unrestricted license to practice medical physics (parity with the requirement for
physicians). However, since few states require licensure for medical physics, an alternate
pathway through letters of attestation has been established.
Lifelong learning and self-assessment require 500 continuing education credits over the 10-year
MOC cycle consistent with the ACR guideline of 50 credits/year.5 The ABR must be responsive
to professional standards active in the industry. Extensive online continuing education (CE)
programs are available through all of the professional societies to offset restrictions in time and
travel budgets. However, to further assist, an optional method to acquire credits has been
established through self-directed educational projects (SDEPs). SDEPs allow prospective
integration of designated professional activities into educational self-improvement projects
providing 15 credits each.
Self-assessment is accomplished through participation in 20 self-assessment modules (SAMs)

over the 10-year cycle. The CME credits obtained for SAMs may be counted as part of the
lifelong learning requirement. SAMs include questions (but no pass/fail) with feedback to the
diplomate.
It is reasonable to expect diplomates to maintain the essentials of core knowledge fundamental to

the practice of medical physics and to remain up-to-date on evolving technologies. Additional
justifications for requiring the single cognitive exam taken toward the end of the 10-year cycle
include: (1) the expectation of the public that such a process be in place for the medical
profession;6 and, (2) Recommendation 7.2 of the Institute of Medicine report on medical errors7
to "implement periodic re-examinations." Every effort will be made to structure the exam as a
productive self-improvement vehicle and not as simply a hurdle to surmount.
The need for MOC in medicine is self-evident and in concert with our sense of professional
responsibility. Carefully considered programs have been instituted by the ABR to accomplish the
task responsive both to our professional realities and to constraints applied by the ABMS. The
requirements are reasonable and can be accomplished by conscientious medical physicists in the
course of fulfilling their practice obligations.
459
AGAINST THE PROPOSITION: Jeffrey Garrett, M.S.
Opening Statement
The American Board of Medical Specialties has mandated that all member boards establish
Maintenance of Certification programs that must follow a general set of guidelines. I contend that
our ABR MOC Trustees should consider that many radiologic physicists, myself included, may
not have the financial and/or staffing resources to fully comply with the current MOC program1
and that the requirements must be appropriate for those being monitored. Additionally, there
should be a logical reason for the requirements that are specified. Parts 1 and 3 are fairly
reasonable and straightforward. Therefore, I will argue against the Proposition based on parts 2
and 4.
The required 500 continuing education credits, average 50 per year, and 20 self-assessment
modules do not seem like a lot at face value. However, bear in mind that there are only 30–35
credits available per major meeting. If unable to attend at least one such meeting every year,
satisfying this CE requirement will be difficult to say the least. Since 2000, I have only been able
to attend 1 AAPM, 1 ASTRO, and 1 Summer School for reasons such as tight hospital budgets
(travel budgets were frozen) and low staffing. Technically, I would be behind schedule at least
130 credits in just 5 years! Radiation oncologists need only 8 SAMs. 8 Anesthesiologists need
only to accumulate 350 credits for the entire component. 9 Internal medicine physicians are only
required to complete 100 credits10—which includes lifelong learning, self-assessment and
assessment of performance in practice! Is it reasonable to place such high demands on physicists
who have much lesser means than physicians?
I suggest reducing the number to something more manageable like 300 total credits. This would
still require frequent attendance at annual meetings yet be flexible enough to allow for years
when hospital budgets are tight or staffing is short. It also provides some incentive to perform
additional research into relevant material rather than taking a quiz simply for the sake of
accumulating credit—unless that is the intent of MOC.
Assessment of performance in practice is currently under review for practically all specialties.
Some have suggested that for physicists this could be a peer review. I see a few problems with
this idea. First, unless you have a friend who is a physicist in close proximity and is willing to do
this for free, you will have to shell out several hundred dollars, maybe even thousands, to satisfy
this requirement. Is that reasonable? Second, multiphysicist institutions present problems as to
who is actually responsible for the program being audited. Finally, having no required standard
review process will lead to widely varying peer reviews. My suggestion for therapy physicists is
to allow Radiologic Physics Center (RPC) TLD results to be submitted. These would
demonstrate that patient care has been administered properly according to established protocols.
The demonstration of proper care could be traced to an individual, and the cost would not come
out of the physicist's own pocket. Similar tests could be used for diagnostic physicists.
Rebuttal: Stephen Thomas, Ph.D.
I am gratified that my colleague finds parts 1 and 3 to be "fairly reasonable and straight forward."
Certainly, I concur and, thus, will focus on addressing his concerns with the other components.
460
The requirement for CE credits is linked directly to the ACR guideline for medical physicists. 5 It
would be incongruous if the MOC program did not support this standard. The issue is how to
obtain those credits with a reasonable expenditure of effort. I understand the constraints on
meeting attendance due to staffing considerations. I am less sympathetic with monetary
arguments in light of salaries commanded by medical physicists today.11 We should be prepared
to expend some personal financial resources in fulfilling professional responsibilities. However,
there are ample means to acquire credits at home through local CME programs and web-based
opportunities. Also, SDEPs are functional and can be designed prospectively into professional
activities and personal education plans. Note that the requirement for category 1 credits is 250
over the 10-year cycle (SDEPs would provide the others), well under the 300 deemed
"manageable."
Practice Performance Evaluation (PPE) (part 4) remains in the process of development. The
intent is to produce structured programs that could identify where improvement in practice might
be incorporated. Currently, the three areas of radiologic physics being considered for PPE
projects are: professional and regulatory guidelines; safety; and educational activities. Peer
review would be included under the first with a working model for therapeutic radiologic physics
published recently by AAPM TG 103 (Ref. 12). The AAPM MOC TG will play a critical role in
shaping part 4 as it evolves and may well evaluate processes such as those suggested by my
colleague.
In summary, the ABR MOC programs are reasonable and promote the integrity and professional
interests of our medical physics discipline.
Rebuttal: Jeffrey Garrett, M.S.
Dr. Thomas states that MOC was not developed inside a vacuum. As proof of this he cites
various conferences held to discuss particulars of the program. Absent, however is any attempt to
collect opinions from physicists who are required to participate in ABR Radiological Physics
MOC. A scan through one of the summaries of these conferences quickly reveals that the
overwhelming emphasis is on physicians—not physicists. Is it reasonable to impose requirements
on physicists that were originally developed for physicians? My hope is that the ABR Trustees
will indeed listen to the recommendations offered by AAPM TG 127.
Dr. Thomas also presents a summary of ways to earn points toward lifelong learning and self-
assessment. Most would agree that the majority of these credits would be earned by attending one
of the professional meetings. As this is not possible every year, Dr. Thomas suggests that we
should spend countless hours either at work or at home working on CE credits via SDEPs and
SAMs. Why not simply make the requirements a little more reasonable? After all, isn't this
maintenance of certification, not certification?
Also, note that SDEPs are prospective. Therefore, the countless hours we have spent
commissioning and testing equipment in the last 3–5 years are not applicable. This is just plain
wrong. Maybe my hospital will purchase a second CyberKnife so I can get SDEP credit. We do
not have the time to sit around and think of projects to do for MOC, much less the time to write
up the project and submit it to one of the MOC Trustees for review and approval.
MOC will have a significant and profound impact on the future of Medical Physics. Why not
make it a positive impact and listen to those who are actually affected by MOC by developing a
program that can be completed by all—not just a select few?
461
REFERENCES
1. The American Board of Radiology Web Site, Radiologic Physicists Maintenance of

Certification requirements (http://www.theabr.org/RP_MOC_Req.htm).
2. J. E. Madewell, "Lifelong learning and the maintenance of certification," J. Am. Coll.
Radiol. 1, 199–203 (2004).
3. S. R. Thomas, W. R. Hendee, and B. R. Paliwal, "The American Board of Radiology
maintenance of certification (MOC) program in radiologic physics," Med. Phys. 32, 263–267
(2005).
4. S. R. Thomas, R. L. Morin, and B. R. Paliwal, "Maintenance of certification in radiologic
physics," J. Am. Coll. Radiol. 3(4), 299 (2006).
5. ACR Practice Guideline for Continuing Medical Education, Revised 2005, Effective
10/01/05 (http://www.acr.org/s_acr/bin.asp?CID=548&DID=12195&DOC=FILE.PDF).
"The role of physician specialty board certification status in the quality movement," J. Am. Med.
Assoc. 292, 1038–1043 (2004).
7. To Err is Human—Building a Safer Health System, edited by L. T. Kohn, J. M. Corrigan,
and M. S. Donaldson (National Academy, Washington, DC, 2000).
8. The American Board of Radiology Web Site, Radiation Oncology Maintenance of
Certification requirements (http://www.theabr.org/RO_MOC_Req.htm).
9. The American Board of Anesthesiology Web Site, Maintenance of Certification in
Anesthesiology (http://home.theaba.org/anes-moc.asp).
10. The American Board of Internal Medicine Web Site, Maintenance of Certification
(http://www.abim.org/moc/index.shtm).
11. AAPM Salary Survey—2005 (AAPM web site: www.aapm.org).
12. P. H. Halvorsen et al., AAPM Task Group 103 Report on Peer Review in Clinical
Radiation Oncology Physics, J. Appl. Clin. Med. Phys. 6(4), 50–64 (2005).
462
10.13. Peer reviews of medical physics practices often yield little

information because the AAPM has not been proactive in
developing appropriate peer-review guidelines
Michael Gossman and Per Halvorsen
OVERVIEW
In a recent Point/Counterpoint, Dr. Stephen Thomas stated that ―…peer review would be
included under the…professional and regulatory guidelines…‖ of the Practice Performance
Evaluation program required for Maintenance of Certification for ABR diplomates. 1 In
anticipation of this, or for other reasons, many medical physicists have begun to organize such
peer reviews. Unfortunately, however, comprehensive guidelines for peer reviews have not been
developed. Lacking such guidelines, some medical physicists have been known to conveniently
―select‖ individuals who will provide ―friendly‖ reviews of their practices. Such peer reviews are
neither useful nor are they suitable for proper practice performance evaluation. It might be
argued that leadership organizations such as the AAPM should have taken a more proactive role
in development and dissemination of such guidelines for peer review, and this is the topic
Arguing for the Proposition is Michael S. Gossman, M.S., DABR. Beginning academically at
Indiana University and the University of Louisville, he received B.S. and M.S. physics degrees,
followed by medical physics education at Vanderbilt University, and certification in Therapeutic
Radiologic Physics by the American Board of Radiology in 2003. He currently serves as a
member of TG-152 and the Therapy Physics Radiation Safety Subcommittee of the AAPM, and
has served as a reviewer for the Medical Dosimetry and Applied Clinical Medical Physics
journals and as a Medical Consultant to the U.S. Nuclear Regulatory Commission. Mr. Gossman
is the Chief Medical Physicist and RSO at the Tri-State Regional Cancer Center in Ashland, KY.
Arguing against the Proposition is Per Halvorsen, M.S., DABR. Mr. Halvorsen is a radiation
oncology physicist at the Middlesex Hospital Cancer Center in Middletown, Connecticut. Since
receiving his MS degree in Radiological Medical Physics from the University of Kentucky in
1990, he has worked in large academic medical centers and private community clinics. He has
been active in the AAPM and the ACR on professional practice issues with particular focus on
practice standards and peer review, serving on the ACR's Radiation Oncology accreditation
committee and chairing or serving on many committees and Task Groups within the AAPM
Professional Council. He is currently an At-Large member of the Board of Directors.
FOR THE PROPOSITION: Michael Gossman, M.S.
Opening Statement
Peer reviews are best accomplished by outsiders, yet medical physicists routinely self-review.
Changes are needed to ensure the effectiveness of peer reviews. The purpose of the peer review
463
process is to provide the incentive and opportunity for medical physicists to improve their
physics quality assurance programs.2,3 The recommendations of an outside qualified medical
physicist can be extremely valuable in establishing policies and procedures, including
implementation of quality assurance guidelines developed by the AAPM and/or patterned
throughout the field. This is also important in order to ensure operational conformance to
regulations governing a medical physics practice.
To maintain high professional performance standards for medical physicists, the AAPM, ACR,
and other organizations have recommended that peer reviews be conducted. 4,5,6,7 Many medical
physicists are still not obtaining peer reviews. A solo medical physicist should obtain a peer
review on an annual basis.2 Where multiple physicists are involved, the chief medical physicist
should obtain a peer review separate from the review for staff physicists. Staff physicists have no
obligations other than those tasks assigned and directed by the chief physicist. Staff medical
physicists should receive an annual performance review by their chief physicist, but the program
as a whole is solely the responsibility of the chief physicist.
The routine of a physics program and the governing policies and procedures for the practice are
strictly the obligation of the physicist in charge. In this sense, the solo physicist is the chief.
Whether acting as consultants or not, the obligations of the physicists are to control and
implement the quality management programs for all facilities in which they operate. For each
facility, peer review analysis should be conducted on the whole program.
As it exists in our field currently, some chief medical physicists assign a staff physicist or part-
time consulting physicist to review the very same program or group with which they work. Even
if a subordinate physicist agrees to take on that responsibility, the peer review process suffers.
This is absolutely not what is meant by ―mutual agreement‖4 for two reasons. First, subordinate
medical physicists provide some of the operational work and would, therefore, be attesting to the
accuracy of their own work. Second, subordinate medical physicists would be commenting on
the program their superior instituted, including opinions on the chief medical physicist's actual
work. This creates significant conflicts of interest.
For any purpose, whether for satisfying site accreditation requirements or for application toward
certification maintenance, this type of review is a ―sham‖! The effectiveness of the peer review
process boils down to the fundamental understanding of its intended purpose: a peer review
should evaluate the entire physics quality assurance program of the chief medical physicist. A
review should be made by an outside qualified medical physicist who specializes in the same
area. That individual should be unaffiliated with, and should not be providing physics work of
any sort for, the facility so as to avoid bias or impropriety. Where appropriate guidance on the
peer review process is lacking, the AAPM should take the initiative and recommend new peer
review procedures that will improve the profession.
AGAINST THE PROPOSITION: Per Halvorsen, M.S.
Opening Statement
Our work as medical physicists can have a significant impact on the health and safety of our
patients. The Institute of Medicine recognized that the continued professional competence of
healthcare staff must be periodically assessed.8 Our professional organizations have recognized
464
the importance of peer review in this context. 2,4,5,7 This shows broad agreement on the value of
peer review in medical physics, and on this point I am sure that Mr. Gossman and I agree.
The debate centers instead on whether ―peer reviews of medical physics practices often yield
little information because the AAPM has not been proactive in the development of appropriate
guidelines.‖ Most medical physics peer reviews do not yield little information. In the rare cases
that this occurs, the villain is not any failure by the AAPM to develop appropriate guidelines.
Rather, the blame lies with the individual medical physicists participating in the review for
failing to consult the many resources available to ensure that the peer review is appropriately
comprehensive and yields a productive critique.
Though no peer-reviewed publication directly addresses the assertion at the center of this debate,
I can relate my personal experience with more than 40 peer reviews. For more than a decade, I
have been a surveyor for the ACR's Radiation Oncology accreditation program. During the same
period, I have performed several privately arranged peer reviews, and have also been reviewed
myself through both mechanisms several times. In all instances, the physicists involved worked
conscientiously to ensure that the review was substantive in scope, appropriate in depth, and
resulted in productive collegial critique.
The AAPM Task Group 103 published guidelines for peer review in clinical radiotherapy
physics,2 and TG 127 is currently working on a peer review model for all medical physics
specialties, in accordance with the ABR's Maintenance of Certification requirement for
Professional Quality Improvement. The ACR maintains an active accreditation program, and
physicists serving as volunteer surveyors gain valuable experience in how to review and critique
a medical physicist's practice. The American Board of Medical Specialties (ABMS) maintains
links on its website to all 24 member boards; many contain useful information about their peer
review programs.
The medical physicist, as an independent professional, should construct a peer review method
that is appropriate for the practice environment by consulting the aforementioned resources and
others as appropriate. I trust that most medical physicists have the intellect and integrity to
exercise their professional judgment on what format would be most appropriate for their
situation. Hopefully, most of us prefer to receive a tangible benefit from the effort in the form of
a useful critique, rather than spending effort on an uninformative review.
In summary, the AAPM is active in developing guidelines and tools for its members to practice
medical physics effectively, including providing guidelines for peer review. We are not a society
of technicians, but of professionals. Each individual professional should have the ability and
integrity to invest the effort to assess the possible formats and appropriate scope of review in
order to design a peer review process that will be productive for the reviewed physicist.
Rebuttal: Michael Gossman, M.S.
The question that remains after these opening remarks is ―how can the AAPM astutely direct the
process of peer reviews?‖ I agree with Mr. Halvorsen that medical physicists should work
conscientiously to ensure that their peer reviews are substantive in scope and appropriate in
depth, resulting in productive collegial critiques. Such critiques cannot be productive when they
are not independent. My colleague cannot dispute that some peer reviews are in fact shams or, at
the very least, are not conducted with the level of independence needed to ensure that they are
unbiased and, hence, constructive.
465
Nevertheless, the debate cannot end here. Neither the ABMS nor the ACR has published
resources that specifically address peer review methods for medical physicists. Although the
ACR suggests that procedures should be in accordance with AAPM guidelines, the current set of
resources, including those made available in TG-11, TG-40, and TG-103, are insufficient and
therefore inappropriate. More comprehensive guidelines that encourage independence will lead
to more productive critiques and foster a higher level of support and quality assurance for
medical physicists at all levels.
I challenge the currently active AAPM Task Group 127 to lead the medical physics community
to a quality peer review process that is complete and broadly supportive of all medical physicists
within our association. I believe the work from this group can produce a report which
encompasses the method of peer review for all specialties of medical physics including solo
practice, as well as for larger facilities where a chief medical physicist oversees work from
subordinate physics staff.
It is my hope that this Point/Counterpoint will provide insight to those members who are
considering participation in this professionally constructive quality assurance process and,
equally, those who are currently involved, all of whom should consider the points mentioned here
as a step toward improving the peer review process.
Rebuttal: Per Halvorsen, M.S.
Mr. Gossman appears to believe that staff medical physicists are merely well paid technicians;
how else to explain the statement that ―Staff physicists have no obligations other than those tasks
assigned and directed by the chief physicist.‖ I strongly disagree. The AAPM-ACMP joint
definition of a Qualified Medical Physicist (QMP)9 states: ―For the purpose of providing clinical
professional services, a Qualified Medical Physicist is an individual who is competent to practice
independently one or more of the subfields of medical physics‖ [emphasis added]. While medical
physicists in training (e.g., residents or junior physicists not yet board certified) may not be
subject to the type of peer review being debated here, all QMPs are capable of practicing
independently and should therefore be expected to contribute their professional judgment and
suggestions for practice improvement, even when serving as staff medical physicists.
Mr. Gossman states: ―…some chief medical physicists assign a staff physicist or part-time
consulting physicist to review the same program or group with which they work…This is
absolutely not what is meant by `mutual agreement'…‖ On this point I agree with Mr. Gossman
and, indeed, the TG-103 report that he references clearly states that ―The reviewer should, as
much as practical, be independent from the reviewed physicist (for example, no business
partnership or close personal relationship).‖
In summary, Mr. Gossman fails to substantiate the proposition that peer reviews of medical
physics practices often yield little information and, furthermore, fails to recognize that our QMP
colleagues are professionals expected to contribute their independent judgment to their practice
environment regardless of their job title. I believe we can, indeed must, meet that higher
standard.
REFERENCES
1. S. R. Thomas and J. A. Garrett, ―The proposed MOC requirements for physicists are
reasonable,‖ Med. Phys. 33, 2319–2322 (2006).
466
2. P. H. Halvorsen, I. J. Das, M. Fraser, D. J. Freedman, R. E. Rice, III, G. S. Ibbott, E. I.

Parsai, T. T. Robin Jr., and B. R. Thomadsen, ―AAPM Task Group 103 report on peer review in
clinical radiation oncology physics,‖ J. Appl. Clin. Med. Phys. 6, 50–64 (2005).
3. G. J. Kutcher, L. Coia, M. Gillin, W. F. Hanson, S. Leibel, R. J. Morton, J. A. Purdy, L.
E. Reinstein, G. K. Svensson, M. Weller, and L. Wingfield, ―Comprehensive QA for radiation
oncology: Report of AAPM Radiation Therapy Committee Task Group 40,‖ Med. Phys. 4, 581–
618 (1994).
4. P. H. Halvorsen, J. F. Dawson, M. W. Fraser, G. S. Ibbott, and B. R. Thomadsen, ―The
solo practice of medical physics in radiation oncology,‖ Task Group 11 of the Professional
Information and Clinical Relations Committee. AAPM Report No. 80 (Medical Physics
Publishing, Madison, WI, 2003).
5. American Board of Medical Specialties. Position statement adopted by the ABMS
Assembly (2000).
6. J. S. Kent and T. A. Mian, ―ACR technical standard for the performance of radiation
oncology physics for external beam therapy,‖ Practice Guidelines and Technical Standards,
edited by N. Detorie, R. L. Morin, and J. M. Hevezi (American College of Radiology, Reston,
VA, 2006), pp. 971–977.
7. J. E. Madewell et al., ―American Board of Radiology: Maintenance of certification,‖
Radiology 234, 17–25 (2005).
8. Institute of Medicine, ―To Err is Human: Building a Safer Health System.‖ Quality of
Health Care in America project. Institute of Medicine, Washington, DC (2000).
9. ―Definition of a Qualified Medical Physicist,‖ AAPM Professional Policy 1D, American
Association of Physicists in Medicine, College Park, MD (2004).
467
10.14. Over the foreseeable future, the growth in technical

complexity of radiation therapy will continue to drive the demand
for more medical physicists
M. Saiful Huq and Jason W. Sohn
OVERVIEW
The technology of radiation therapy, including 3D and conformal treatment planning, IMRT,
gamma-knife and tomotherapy, and seed, intravascular and HDR brachytherapy, has created a
nationwide demand for medical physicists. Whether this surge in technology, and the consequent
need for physicists, will continue into the future is the subject of much speculation. This
speculation is the subject of this month's Point/Counterpoint.
Arguing for the Proposition is M. Saiful Huq, Ph.D. Dr. Huq received his Ph.D. in physics from
the College of William and Mary in 1984. He is currently Director of Medical Physics at the
Department of Radiation Oncology of the University of Pittsburgh Medical Center Cancer
Centers. Dr. Huq has served on many AAPM Task Groups and Committees, including TG-51,
and currently is chairing TG-100. He is Chair of the Calibration Laboratory Accreditation
Subcommittee. Dr. Huq is a fellow of the AAPM, a fellow of the Institute of Physics, has been a
co-recipient of the Farrington Daniels award, and has published more than 135 peer-reviewed
journal articles, book chapters, proceedings papers and abstracts.
Arguing against the Proposition is Jason W. Sohn, Ph.D. Dr. Sohn is an Associate Professor of
Radiation Oncology at Case Western Reserve University, and serves as Associate Director of
Medical Physics and Dosimetry. He is a former AAPM chapter president, current member of
several AAPM committees, and an ABR guest examiner. His editorial activities include serving
as a reviewer for the International Journal of Radiation Oncology Biology Physics, Medical
Physics, and AAPM reports. He teaches physics to medical and physics residents. His research
interests are stereotactic radiosurgery, optical and radiological imaging, and intensity modulated
radiation therapy.
FOR THE PROPOSITION: M. Saiful Huq, Ph.D.
Opening Statement
In the past two decades, rapid advances in technology have pushed the development of
technology-intensive, image-guided modalities for cancer treatment that we recognize as modern
radiotherapy. Biologic information from various physiologic imaging modalities is used
increasingly to delineate target volumes accurately, and is becoming an integral part of the
treatment design process. Some of the newer technologies and associated complex treatment
procedures include, but are not limited to, image registration and fusion from various imaging
modalities. Examples include high dose rate, prostate seed and coronary vascular brachytherapy,
and stereotactic radiosurgery and radiotherapy. Image-guided radiotherapy and IMRT require
468
image guidance combined with immobilization devices such as breathing control or gating
devices to minimize the impact of geometric uncertainties of organ motion and setup error.
Accelerators are being marketed with integrated imaging devices to provide a means of seamless
target identification, real time monitoring, delivery modification, delivery verification, dose
reconstruction, and adaptive radiotherapy. 1,2 Patients and their family members are demanding
better and higher quality care.
Clinical implementation of complex computer-augmented technology-intensive devices and

procedures, including dose calculation and optimization algorithms, requires a clear
understanding and critical evaluation of their limitations and performance characteristics.
Because of their background in physical sciences, qualified medical physicists (QMPs) are in a
unique position to evaluate and manage these devices and procedures on an ongoing basis;
perform acceptance testing and commissioning of the software and hardware used for treatment
planning and delivery; establish baseline QA parameters for ongoing safe use of the devices;
develop procedures that satisfy regulatory mandates; design and perform appropriate patient-
specific QA tests to verify correct and safe delivery of intended treatments; evaluate clinical
outcomes of using the new technologies; and, most importantly, act as a technical resource to
educate the radiotherapy community about the safe, efficient and intended use of the
technologies. These procedures are very labor intensive and place a heavy demand on medical
physicists.
The 1994 American College of Radiology survey3 on radiation oncology in the U.S. shows that
between 1983 and 1994 the number of FTE physicists increased by 60%. The 2003 Abt Study of
Medical Physicist Work Values for Radiation Oncology Physics Services4,5 found that between
1995 and 2003, practices offering remote afterloading brachytherapy increased from 46 to 66
percent, multileaf collimators increased from 19 to 79 percent, and electronic portal imaging rose
from 20 to 53 percent. Technology-intensive procedures such as prostate seed brachytherapy rose
to 89 percent, 3D conformal therapy (non-IMRT) to 92 percent, coronary vascular brachytherapy
to 74 percent, and record and verify systems to 87 percent. The median relative work estimates
for a QMP increased by a factor of 12 for IMRT treatment planning, 14 for special medical
physics consultation and 16 for IMRT special physics consultation.5
These numbers indicate that the discipline of radiation oncology is continually changing in
response to technology, practice, and state and federal regulations. They are a testament to the
vital role that medical physicists play in radiation oncology and the delivery of patient care. As
the growth in image-guided technology escalates, treatment procedures increase in complexity,
and regulatory agencies place a premium on efforts to enhance safety, the demands for medical
physicists will continue to grow. Under these circumstances, medical physicists will become
even more important in the delivery of optimum patient care.
Rebuttal
Dr. Sohn supports his position with the assertion that once "technologies are in place"
computerization and automation of physics duties and streamlining of new tasks for emerging
systems "as experience is acquired" will likely reduce the demand for physicists. He also states
that "improved clinical results from the use of advanced technology" is increasing clinical
workloads. Most of these statements are consistent with my opening statement that clinical
implementation of technology-intensive devices and procedures puts a heavy demand on medical
physicists. I do not agree, however, with his contention that streamlining of physics activities will
decrease the demand for physicists. Streamlining is necessary to ensure ongoing safe use of
469
various devices and procedures, and to keep pace with the ever-growing patient-specific
multitude of procedures that are the product of emerging technology. Instead of reducing the
workload, streamlining enables task accomplishment in a timely fashion. It is a myth that
computerization and automation reduce workload.
Citing the helical tomotherapy machine as an example, Dr. Sohn argues that convergence of
multiple devices will give rise to less complicated systems which will reduce the demand for
physicists. To the contrary, the helical tomotherapy machine is a very complicated technology
that not only performs most of the functions of the component devices that Dr. Sohn has
identified, but also offers additional complex state-of-the-art treatment planning and verification
procedures such as setup verification, delivery modification, delivery verification, and dose
reconstruction. It also enables adaptive radiotherapy. This is a complicated technology that
requires careful evaluation by medical physicists before safe implementation in the clinic.
Clearly, convergence of multiple devices does not necessarily lead to a reduction of workload.
Dr. Sohn argues that the cost of physics support for new technologies should be balanced by
keeping the number of physicists constant. Otherwise, as he states, "they will price themselves
out of the market." He does not provide any rational basis for making such an argument. Image-
guided radiation therapy will continue to grow and so will the demand for qualified medical
physicists.
AGAINST THE PROPOSITION: Jason W. Sohn, Ph.D.
Opening Statement
Over the past two decades, few medical fields have been more technologically progressive than
radiation therapy. Clinical workloads in radiation therapy have increased because of greater
numbers of patients and increased complexity of treatment procedures. The upward trend in
patients is not only a result of an aging population, but also a product of improved clinical results
from the use of advanced technology. The workload expansion will certainly continue to
increase, and investments to improve reliability and ease of use for clinical systems will be
rewarded by the market.
The two trends of advancing technology and increasing demand for services ultimately are likely
to reduce the demand for physicists, in order to make treatment costs affordable. The service
cost/physics manpower issue has been studied by ACMP and AAPM.4,6 Physicists' service-
related charges have increased by 30%. When a strong demand for the delivery of services hits a
personnel shortage, the impasse is generally eased by computerization and automation. I have
witnessed the progression of medical physics from the simple mechanized delivery of a few
select energies and fields for simple dose plans, to computerized systems directed by highly
evolved control systems that deliver complex dose treatments. The entire clinical arena, from
dosimetry to delivery, is trending toward self-contained computerized systems. No area of
radiation therapy is untouched by significant advances in imaging technology, accurate patient
positioning, improved computer modeling of radiation treatments, and precision-controlled
delivery systems. All of this points to an important trend: the forced convergence and maturation
of an industry to meet growing technology-focused demands.
Maturation of a process occurs when new tasks performed by a physicist for emerging systems
eventually become simplified and streamlined as experience is acquired. Frequently a majority of
470
a physicist's duties for new technologies can be mechanized and automated once the technologies
are in place. Clinics are beginning to desire entire systems that require little maintenance and
fewer personnel, like the evolution of automobiles. A good example is the modern linear
accelerator. We used to perform quality checks once a week. Today, greater reliability has
reduced these quality checks to once per month.
As clinical systems mature, they become subject to convergence. Multiple devices, even entire
diagnostic and treatment systems, are being folded into singular devices. Virtual simulation using
CT scanners is steadily replacing x-ray simulators as clinics invest in a single diagnostic machine
to replace two imaging systems. X-ray simulators may soon join cobalt teletherapy machines as
functional curiosities from the past. The helical tomotherapy system is an excellent example of
convergence. Should this technology prove itself in the clinic, one machine may replace the CT
scanner, simulator, linac, and accompanying portal imaging system. Converging treatment
systems will be less complicated, require less space, and will eventually reduce the demand on
physicists.
We are now at the stage where we can determine whether new, advanced treatment
systems/methods, including IMRT and IGRT, improve clinical outcomes. Furthermore, the
improvement must be significant enough to justify the increased treatment cost. I believe that the
demand for physicists will stay fairly constant as technology improves to extract greater
efficiency from current and emerging clinical systems. Otherwise, the cost of physics support for
new technologies will be so expensive that they will price themselves out of the market.
Rebuttal
As Dr. Huq points out, the demand for physicists has been increasing over the past several years.
That is why I assert that market forces will move to automation to prevent crippling cost
increases related to continued growth of physics personnel. Clinics will weigh the cost of hiring
more physicists with that of buying increasingly closed, automated systems. Even now, most of
the systems cited as "growing" in importance are automated replacements of older, labor
intensive techniques: MLC replacing blocks, electronic portal imagers replacing film, HDR
replacing LDR, Record and Verify replacing paper charts are examples. IMRT procedures have
become routine. In the beginning, complex quality assurance measurements for each patient and
plan required substantial time and effort from physicists. Now in many institutions, physics
assistants are performing QA under a physicist's supervision. The AAPM recognizes the
responsibilities of physics assistants and dosimetrists.7 Complex dosimetry verifications are
being simplified and reimbursement has been lowered as the IMRT process has become mature
and streamlined.
Technology is a limitless enterprise, able to provide increasingly sophisticated techniques so long

as they are affordable. We cannot continue to expand the clinical demand for physicists (and the
corresponding rise in healthcare costs) without insurers stepping in to demand relief. The system
is self-regulating. It is now time to shift away from personnel-intensive technologies to
automated technologies in radiation oncology.
REFERENCES
1. D. A. Jaffray, J. H. Siewerdsen, J. W. Wong, and A. V. Martinez, "Flat-panel cone beam

computed tomography for image guided radiation therapy," Int. J. Radiat. Oncol., Biol., Phys.
53, 1337–1349 (2002).
471
2. T.R. Mackie, G. Olivera, G. Fang, J. Kapatoes, E. Fitchard, P. Reckwerdt, D. Shepard, K.

Ruchala, J. Balog, and S. Aldridge, "Helical tomotherapy: soon to be an option for the delivery
of intensity-modulated radiation therapy," 4th Int. Symp. on 3D Conformal Radiation Therapy
and Intensity Modulated Radiation Therapy in the New Millennium (Houston, TX).
3. J. B. Owen, L. R. Coia, and G. Hanks, "The structure of radiation oncology in the United
States," Int. J. Radiat. Oncol., Biol., Phys. 39, 179–185 (1997).
4. Abt Associates Inc., "The Abt Study for Medical Physicist Work Values for Radiation
Oncology Physics Services," prepared for the American College of Medical Physics (ACMP)
and the American Association of Physicists in Medicine (AAPM), October 3, 1995.
5. Abt Associates Inc., "The Abt Study of Medical Physicist Work Values for Radiation
Oncology Physics Services," prepared for the American College of Medical Physics (ACMP)
and the American Association of Physicists in Medicine (AAPM), June, 2003.
6. ACMP Task Group Report: Survey of Physics Resources for Radiation Oncology Special
Procedures (American College of Medical Physics, 1998).
7. AAPM Report Number 38: The Role of a Physicist in Radiation Oncology (American
Association of Physicists in Medicine, 1993).
472
10.15. The future will not need clinical therapy physicists

George Starkschall and George W. Sherouse
OVERVIEW
Recently William Joy, founder of Sun Microsystems, published an article in Wired entitled "The
Future Will Not Need Us." This provocative article caused the editor of the Point/Counterpoint
series to think about whether there is a future role for medical physicists in radiation oncology (or
in any discipline, for that matter). With the rapid evolution of nanotechnologies, microrobotics
and intelligent machines, it does not require much imagination to foresee that many (if not all) of
the functions presently provided by clinical therapy physicists will ultimately be attainable by
machines, probably with greater uniformity, reproducibility and cost-effectiveness. Some would
argue, however, that medical physics requires clinical adjustments and judgments that will
always be beyond the reach of automated systems. This controversy is the subject of this issue of
Point/Counterpoint.
Arguing for the Proposition is George Starkschall, Ph.D. Dr. Starkschall is Associate Professor of
Radiation Physics at the University of Texas M. D. Anderson Cancer Center. He has served on
many AAPM committees including Radiation Therapy, Program, Awards and Honors, and
Development. He also served on the Board of Directors from 1992 to 1994. He was Scientific
Director of the 1996 AAPM Annual Meeting and has organized several national and international
symposia. He has authored or co-authored articles on radiation treatment planning, radiation
oncology imaging, and electron-beam dose calculations, and edited several books on quality
assurance and conformal therapy. He is certified in Therapeutic Radiological Physics by the
ABR and in Radiation Oncology Physics by the ABMP.
Arguing against the Proposition is George W. Sherouse, Ph.D. Dr. Sherouse, a Florida native,
received his M.S. in medical physics and clinical training at the University of Florida. After a
brief stint as a product developer with a Canadian treatment planning company, he was recruited
by the University of North Carolina to help reinvent radiotherapy computing. The result was a
widely distributed 3D treatment design system called GRATIS TM, which included the Ph.D.
winning Virtual SimulatorTM.1 Dr. Sherouse passed the 1990s on the faculties of Duke University
and the Medical University of South Carolina, and is currently self-employed as a consulting
medical physicist and mercenary product developer.
FOR THE PROPOSITION: George Starkschall, Ph.D.
Opening Statement
The need for clinical physicists is two-fold: (1) Radiation doses presently delivered in radiation
therapy are high enough to cause unacceptable damage to uninvolved tissue, and (2) there is no
direct indication of where the radiation dose is delivered. Consequently, validation is needed that
the patient dose is correct and is delivered to the correct location. To validate dose delivery,
medical physicists commission a treatment machine by acquiring data characterizing the
473
radiation output. The machine is subjected to quality assurance procedures to verify that the
output does not change with time. Treatment planning systems are used to model radiation
interactions in patients and to calculate where the radiation dose is delivered. Portal imaging is
used to verify the accuracy of delivery. All of these tasks have significant medical physicist
involvement. If radiation doses were significantly lower, or if dose delivery could be determined
with greater confidence, the role of the clinical physicist in radiation oncology could be reduced.
In the future, cancer therapy is likely to remove one, or both, of the needs for clinical physicists.
Future therapies will rely less on radiation than they do today. Future therapies will target cancer
on a molecular level, and will likely be less toxic than present systemic therapies. One example is
molecular chemotherapy that targets critical events in the carcinogenic process, such as growth
factors, angiogenesis, and immune response. Another example is gene replacement therapy in
which dysfunctional tumor suppressor genes are replaced. A third example is cytotoxic gene
therapy in which tumor cells are targeted by toxic genes. Yet another example is the use of
angiogenesis inhibitors, such as angiostatin and endostatin, to prevent proliferation of tumors by
blocking generation of blood vessels. Finally, immunotherapies such as vaccine and antibody
therapy may evolve to reactivate host anti-tumor activity. Common to all of these therapies is the
use of targeted molecules or cells, with a reduction in systemic toxicity. Radiation, if used at all
in the future, will be a supplemental technique for the newer molecular therapies. Radiation
doses are likely to be significantly lower, and the planning and verification of radiation
treatments will not be so critical.
One source of the uncertainty in dose delivery is the uncertainty in output of the radiation source.
This uncertainty results in extensive commissioning measurements and an elaborate quality
assurance program. Accelerators are being manufactured today, however, to deliver doses with
much greater reliability. Consequently, commissioning these new machines may require little
more than spot checks. In addition to a reliable source output, the dose delivered to the patient
will also be more predictable. Electronic portal imaging devices are being developed that can
acquire quantitative dose information in real time. This information can be compared with the
dose distribution predicted from the treatment plan, and the difference fed back to the machine
for real-time compensation. The use of such feedback mechanisms will allow us to "treat by
wire" much as we now "fly by wire" in sophisticated aircraft.
This brave new world is not in the immediate future. Short-range job prospects for clinical
physicists are excellent. In the longer-range future, however, as the use of radiation for cancer
treatment declines and the sophistication of treatment machines increases, clinical physicists may
go the route of chimney sweeps and blacksmiths.
Rebuttal
Dr. Sherouse justifies the future need for clinical radiation therapy physicists on the assertion that
medical physicists provide the required expertise to ensure ongoing safety and quality in patient
care. My argument is that the tasks performed by physicists to fulfill these needs may no longer
be necessary. Commissioning and quality assurance procedures are presently performed on
radiation equipment because we do not have the confidence that the output of a machine is
predictable and unchanging. When we can predict the output of a radiation machine with a high
degree of confidence, when we can demonstrate that the output of the machine remains constant
with time, and when we can determine with confidence the true radiation dose received by the
patient, our clinical justification will be severely curtailed.
474
Disasters such as the Therac 25 incident that Dr. Sherouse mentioned will, unfortunately,
continue to occur. It is unlikely that the software failure that led to the incident would have been
uncovered during any routine machine commissioning, quality assurance, treatment planning, or
treatment verification procedure. It would be very difficult to justify supporting a full-time
clinical physicist with little else to do than to wait for such an incident to happen. The actions of
the on-site clinical medical physicist did not prevent the Therac 25 incident from happening; his
role was to respond to the emergency and determine its cause. We learn from our experience,
and, in particular, from our failures. It is highly unlikely that future linear accelerators will be
designed with software that will fail in precisely the same manner as the Therac 25. The product
has been improved by that tragic incident. Fortunately, such incidents are rare.
The role of the clinical medical physicist is likely to evolve from that of radiation consultant,
playing an essential role in the planning and delivery of radiation treatment to that of radiation
fireman, responding to critical emergency events. That is a very changed role for the medical
physicist, and it is not the clinical role we know today.
How will clinical medical physics be supported in the future? Present reimbursement schemes
recognize the clinical role of the medical physicist in supporting quality assurance directing,
treatment planning, and providing consultation. It is not likely that health care payers would be
willing to support one or more individuals at each radiation therapy facility whose sole purpose is
to prevent the unforeseeable incident from occurring or to respond to critical events. Nor is it
likely that creative, talented scientists would be attracted to a profession whose image would be
similar to that of the Maytag repairman who sits around waiting for the emergency call that never
comes.
AGAINST THE PROPOSITION: George W. Sherouse, Ph.D.
Opening Statement
Bill Joy's disturbing article contemplates the near-term potential for technological developments
in self-replicating systems to render humanity as we know it obsolete. This could manifest as the
creation of intentionally engineered "improved" physical forms for humans or, just as easily, the
catastrophic proliferation of a lethal invention. In that light the current proposition becomes, "The
future will not need people so neither will it need clinical therapy physicists." Even I can't argue
with that. Dr. Starkschall wins a hollow victory.
But I accepted this assignment because I think there is a more immediate formulation of the
proposition that does bear our attention, namely that "Technological developments in radiation
oncology will, in the near term, render medical physicists unnecessary to effective clinical
practice."
This is a timely conversation. Zealous advocates of the emerging generation of chattering teeth
treatment units are already claiming that their devices will be self-contained, self-calibrating,
self-monitoring and generally responsible for their own behavior. The caveat "except when they
fail" is rarely spoken. Our history provides numerous cautions against such overconfidence in
technology.2 The well-documented story of the Therac-25 incidents3 alone ought to maintain our
sobriety. Three patients were killed by machine faults that the vendor insisted could not happen.
It was left to a clinical medical physicist to prove otherwise.
475
It is the very nature of life-critical technology that increasing complexity requires more, not less,
expert attention. That fact exists in tension with Arthur C. Clarke's famous observation that any
sufficiently advanced technology is indistinguishable from magic. The risk exists that at the same
time technology becomes more opaque, the dreamy spell cast by the new magic will enable the
displacement of qualified magicians by sorcerers' apprentices. The difference will likely only be
revealed in crisis.
The role of the clinical physicist is, simply, to ask questions and find answers that assure the
ongoing safety and quality of patient care. The qualities that make us uniquely good at that work
are a broad and deep base of interdisciplinary knowledge, formally-trained reasoning skills,
strong curiosity, healthy skepticism and a relentless need to trace any mystery to its source. I
have no fear of being forced into early retirement by a machine outperforming me on these
counts.
In a more perfect world the need for qualified medical physicists would be codified in both the
law and the fiscal structure. In our world, what little law there is requiring clinical physics
services exists largely in reaction to serious accidents and, as evidenced by the latest HCFA rules,
clinical physics has little standing among billable services. As a profession we have spectacularly
failed to police our own training and credentialing in a credible way and we have refused to soil
our hands with the actual hardball politics and business of health care, preferring to bicker among
ourselves about points of ego while others cast our fate. What outcome did we expect?
I see a future that will need, but not acknowledge that it needs, qualified clinical medical
physicists. That is a future in which we all lose, providers and patients alike. Advancing
technology may briefly serve as a convenient smokescreen but the responsibility will rest
squarely with a profession too self-absorbed to bother defining and justifying its own crucial
existence.
Rebuttal
One can, without a doubt, hear the big Cure For Cancer clock ticking. I'm betting on the
nanomachines myself. But for the indefinite meantime
Dr. Starkschall's statement contains the very kernel of the problem which most concerns me. He
describes the duties of a technician who monitors the performance of radiation machines and
labels that person's job "clinical physicist." My mentors taught me that a clinical physicist is a
health care professional of extraordinarily broad expertise who, first and foremost, takes care of
people. The distinction is fundamental.
Yes, clinical radiation oncology physicists do take care of machines, but that increasingly
demanding task is only one bullet in the job description. Our essential role is vigilance and
strategic intervention in the customized high-tech care of individual people. There is just so much
more to being an effective clinician than making radiation field measurements. It is our failure to
clearly articulate and codify that distinction, among ourselves and in our interface to the world,
which puts us at risk of obsolescence. Even in Dr. Starkschall's future where those dangerous
radiation beams are no longer part of the Cure For Cancer, chances are good that there will still
be an important role for physicist-clinicians capable of harnessing magic-seeming technology
into the service of their patients.
476
And for the record, I have personally employed both a chimney sweep and a blacksmith within
the last two years to provide services for which they were highly trained and uniquely qualified.
REFERENCES
1. GRATISTM and Virtual SimulatorTM are trademarks of Sherouse Systems, Inc.

2. See, for instance, Lessons Learned From Accidental Exposures in Radiotherapy, IAEA
Safety Reports Series No. 17 (International Atomic Energy Agency, 2000).
3. N. G. Leveson and C. S. Turner, "An Investigation of the Therac-25 Accidents,"
Computer 26, 18–41 (1993).
477
10.16. Noncompete clauses in employment contracts violate a

physicist's freedom to practice his/her profession
Shirish K. Jani and Prakash Shrivastava
OVERVIEW
Increasingly, physicists are being asked by their current or potential employer to sign
noncompete clauses promising to not work in the community for a specific period if they leave
the employer. At times these clauses are accompanied by financial incentives to sign. Most
physicists sign these clauses because they want the job or the incentive, or because they do not
plan to leave the employer. But some believe that signing a noncompete clause violates their
fundamental rights and freedoms as an American and as a professional. This controversy is the
subject of this month's Point/Counterpoint.
Arguing for the Proposition is Shirish K. Jani, Ph.D. Dr. Jani received his Ph.D. in molecular
physics from North Texas State University in 1980. After completing a post-doctoral fellowship
at the Medical College of Virginia, he joined the University of Iowa School of Medicine and later
became the director of clinical physics. Since 1993, he has been at the Scripps Clinic in La Jolla,
CA. He is certified by the ABR and the ABMP. Dr. Jani has served on many AAPM task groups
and is an active member of the ACR Commission on Physics. He serves as an ABR Oral
Examiner. Dr. Jani is a Fellow of the American College of Radiology and the AAPM.
Arguing against the Proposition is Prakash Shrivastava Ph.D., MPM(H). Dr. Shrivastava is a
Professor of Radiation Oncology, Professor of Radiology and Professor of Biomedical
Engineering at the University of Southern California, Los Angeles, CA. He serves as the Chief
Medical Physicist and Radiation Safety Officer at the Los Angeles County Medical Center. He
has a Master of Public Management (Health) degree from the Carnegie Mellon University in
Pittsburgh, PA.
FOR THE PROPOSITION: Shirish Jani, Ph.D.
Opening Statement
A noncompete clause in an employment contract may require that the physicist, after leaving
his/her current employment, does not work in the same geographic location for a specified
interval of time. I have two major objections to such a clause.
First, some noncompete clauses are intended to protect trade secrets and the intellectual property
of a company. For example, suppose an employee of a computer software company (such as
google.com) joins the main competition (such as yahoo.com). He/she may possess and could
transfer valuable proprietary information to the new company. In this instance, a noncompete
clause may serve a useful purpose. However, a medical physicist rarely possesses trade secrets or
intellectual property that is important to a new employer. Patients do not switch hospitals or
healthcare providers in order to continue receiving the same medical physics services! Moreover,
478
the intellectual foundation of medical physics activities already exists in scientific journals and
text books. Therefore, a noncompete clause is not needed to protect employers.
Second, a noncompete clause restrains medical physicists from engaging in lawful activities. For
this reason, noncompete clauses are illegal in California. While there are narrow exceptions, the
policy (California Business and Reference Code Section 16600) states: "Every contract by which
anyone is restrained from engaging in lawful activity in a lawful profession, trade or business of
any kind is to that extent void."1
Physicists typically sign contracts with noncompete clauses because (1) they want the job, (2)
they can use financial incentives, and (3) they naively believe that they will never leave the
hospital or institution. The most frequent reason to leave an employer is to seek a better work
environment. Physicists who do not find their current work environment conducive to good
practice often wish to find a new job without moving their residence. The noncompete clause
prohibits them from accepting a nearby job. This is simply overly restrictive and unjust.
The noncompete clause is fundamentally wrong, should be illegal, and is not applicable to our
profession.
Rebuttal
In his position against the proposition, Dr. Shrivastava tries to convince the reader that a
noncompete clause "can be fair, legal and beneficial to both the employer and the employee." Let
us first address the issue of fairness. Strictly enforced, a noncompete provision could mean that a
former employee cannot make a living in the field without relocating. This is not fair. In addition,
employees often leave due to workplace harassment and/or a power struggle. The severity of
these conditions determines how fast an employer loses good employees. It would be far more
beneficial for employers who are unable to retain good staff to identify the cause of losing their
staff and to take corrective action, rather than to resort to a noncompete clause.
Another aspect of fairness deals with employers protecting intellectual property. This protection
can be accomplished by a nondisclosure agreement instead of a noncompete agreement. Such
nondisclosure agreements are valid most everywhere in the United States.
Let us now address the issue of legality. As mentioned earlier, the noncompete clause is illegal in
California. In states where it is legal, the intent is to protect employers but also not to deprive
employees of earning a living in their chosen field. This is a balancing act that hinges on fairness.
An example is the case of Walia vs. Aetna, Inc., No. 992768, San Francisco Superior Court. In
1996, Aetna, Inc., merged with U.S. Healthcare. Aetna required U.S. Healthcare employees to
sign an agreement that prohibited them from working for a competitor in the same state for two
years, or for any competitor for six months, after employment with the company had ended.
Those who refused to sign the agreement were terminated. One of the employees who refused to
sign sued Aetna alleging wrongful termination. Aetna defended by arguing that public policy
would be violated only if Aetna attempted to enforce the agreement. The jury sided with the
employee and resolved that Aetna knew the agreement was illegal and nonetheless required
employees to sign it.
Dr. Shrivastava ends his argument by stating that noncompete clauses "are as American as other
types of negotiated contracts in our society." I completely disagree. In a free market such as ours,
many of the glamorous start-ups are the products of free spirits. Imagine what the computer
479
industry of this country would look like today if IBM had elected to impose, and been allowed to
enforce, noncompete clauses in every possible instance. Competition is American; prohibiting
competition is not.
In summary, when a physicist signs a contract that includes a noncompete clause, typically he/she
is agreeing not to engage in any business of a similar nature in that area after leaving the
employer. It is wise to think twice before signing such a contract.
AGAINST THE PROPOSITION: Prakash Shrivastava, Ph.D.
Opening Statement
A noncompete covenant is a clause in an employment agreement or a separate contract standing

by itself.2,3,4 A number of questions arise when employees are asked to sign a noncompete
covenant as a condition for employment. Is it fair? Is it legal? Is it enforceable? Should I sign it?
Is it fair? To employers who maintain their business or competitive edge by using proprietary,
innovative ideas or technology it is critical that their intellectual property not be taken away by an
employee and delivered to a competitor. It is the employer's right to protect intellectual property
in order to survive in a free business environment. It is reasonable for an employer to protect
itself by preventing the employee from taking knowledge acquired on-the-job to compete against
the employer in a nearby enterprise.
Is it legal? It is in some states and not in others. In California, for example, noncompete
covenants are not legal, but in Ohio they are. States base their laws on principles of fairness to
businesses and also to individuals. An agreement cannot overly restrict employees' rights so as to
violate their freedom to pursue their professions. Some specific limits on time (6 to 24 months)
and distance are often allowed in noncompete covenants, so long as they are applicable to all
employees and do not violate community interests.
Is it enforceable? Even if a covenant is legal and signed by an employee, it may or may not be
enforceable, depending on the conditions of the agreement. Employers may have to prove in
court that the covenant is necessary to protect their legitimate business interests. Not all
employers would be able to convince a court that their business interests are important enough to
prevent employees from working for someone else. If the competition is next door or in close
proximity, however, and if it competes for the same clients, an employer may have a convincing
argument.
Should I sign it? This is where the individual needs legal advice. The answer may depend on
whether you are a new employee, or have been employed for some time without such an
agreement. Mid-employment noncompete agreements may be less enforceable than agreements
for new employees, unless the employee is given some additional advantage by signing them. In
case of new employment, to sign or not to sign is your choice. Before you sign away your
freedom to work for someone else in the same geographic area, make sure that the restriction is
worth the benefits of employment. If you have options, consider working somewhere else. If you
have plans to work in the area for yourself or with someone else, you should negotiate the
noncompete contract before you sign it rather than when you are ready to end it.
480
In summary, noncompete covenants can be fair, legal and beneficial to both the employer and the
employee. How can it be beneficial to the employee? Suppose two physicists work at a facility
where the junior person learns the trade, and then leaves to help a competitor establish a new
facility next door, which takes away all the business. The senior physicist is now out of a job.
There is no single, simple answer to fairness, legality or validity of noncompete clauses. They are
as American as other types of negotiated contracts in our society.
Rebuttal
Dr. Jani concludes that a noncompete clause is fundamentally wrong, should be illegal, and in
any case should not be applicable to our profession as medical physicists.
Right or wrong is not the issue here. An individual's legal rights, and the protection of business
investment or intellectual property are of concern. Legality is a community standard established
to balance business interests and individual freedoms. In some communities with a high density
of competing businesses, noncompete clauses are acceptable, useful, legal, negotiated contracts.
As Dr. Jani correctly points out, these clauses mainly limit the time period within which the
employee cannot compete or help the employer's competitors within a specified region. These
limitations are restrictive but not prohibitive to the extent that the employees cannot earn a living.
Dr. Jani states that noncompete clauses should not apply to medical physicists, although they
could apply to other professionals, because physicists, unlike others, do not have trade secrets or
intellectual property. This surely is not a supportable fact. In addition, applying the clause to
some professionals and not to others would be considered discriminatory in any community. In
fact, in legal disputes courts often check to determine if employers apply the clause uniformly
across the board.
When asked to sign a noncompete clause, a physicist has two options: (1) not take the job or (2)
ask for specific names of regional competitors and agree not to take a new job with them for the
specified period.
REFERENCES
1. California Codes: Business and Professions Code, Division 7: General business

regulations, Part 2: Preservation and regulation of competition, Chapter 1: Contracts in restraint
of trade. (2003), http://www.leginfo.ca.gov/calaw.html.
2. Internet Legal Resource Guide, ilrg.com.
3. Physicians News Digest, Jan 2001.
4. NUPP Legal, Nupplegal.com.
481
10.17. Medical physicists should seek patent protection for new ideas
before publishing articles about them
L. E. Antonuk and Perry Sprawls
Reproduced from Medical Physics, Vol. 26, No. 11, pp. 2220-2222, November 1999
OVERVIEW
Knowledge about a subject grows as research results accumulate about the subject. Some
scientists believe they should publish results quickly in order to stimulate the growth of new
knowledge. In their view, rapid publication of results is an obligation, especially when the results
are from research supported by public funds. Other scientists feel they should protect their results
by patent applications, even though filing such applications delays publication of results. They
claim that they deserve to share in profits from the fruits of their labors, and also that society
benefits because companies will invest in results only when they are protected by patents. The
controversy is becoming increasingly polarized as science becomes more secular and as
scientists, including medical physicists, struggle to identify ways to support research. In this
issue of Point/Counterpoint, two experienced medical physicists explore this polarization.
Arguing for the Proposition is Larry E. Antonuk, Ph.D. Dr. Antonuk, a Canadian citizen,
received his B.Sc. (Physics, 1975) from the University of Calgary and his Ph.D. (Nuclear
Physics, 1981) from the University of Alberta, having worked at TRIUMF in Vancouver. From
1981–1984 he was a Research Fellow for the University of South Carolina working at the
Universite´ de Neuchatel, Switzerland and at the S.I.N. accelerator. From 1984–1987 he was a
Research Associate for the University of Alberta working at the Laboratoire National Saturne
accelerator in Saclay, France. He joined the Department of Radiation Oncology at the University
of Michigan in 1987 where he is presently an Associate Professor of Radiation Physics and
heads the active matrix flat-panel imaging group.
Arguing against the Proposition is Perry Sprawls, Ph.D. Dr. Sprawls received his Ph.D. degree
from Clemson University in 1968 after joining the Emory University faculty in 1960. He is
Professor of Radiology and Radiation Oncology at Emory and served as Director of the Division
of Radiological Sciences. He is on the faculty of several other international universities and is a
Director of the College of Medical Physics, International Center for Theoretical Physics, Trieste,
Italy. He is certified by the American Board of Radiology in diagnostic physics, the American
Board of Medical Physics in diagnostic imaging physics and magnetic resonance imaging and
has served as an examiner for both boards. He is author of a series of textbooks on the physics of
medical imaging.
FOR THE PROPOSITION: L. E. Antonuk, Ph.D
Opening Statement
The rapid and thorough dissemination of new knowledge is widely regarded as among the
highest objectives of those involved in the pursuit of scientific discovery. It is also generally
recognized that the successful translation of laboratory findings into practical application is of
482
critical importance to society at large, especially in light of the heavy dependence on federal
funding of basic research in the U.S. Accomplishing this second goal often requires the
involvement of commercial interests that are willing and able to invest the necessary resources to
transform scientific discoveries or inventions into useful products. However, bringing a new
technology to market is frequently a high-risk endeavor that is unlikely to bring substantial
returns for many years. For this reason, the availability of patent protection through licensing can
be of pivotal importance in the decision of a company to pursue the development of a new
technology. This is especially true for small companies whose success may vitally depend on
some degree of temporary relief from competitive pressures as afforded through licensing of
patents. Moreover, small companies are often considerably more inclined to assume the higher
risks and relatively lower short and medium-term rewards associated with bringing a new
technology to market. Thus, seeking patent protection for new ideas prior to publishing may well
be the determining factor in whether the results of research ultimately benefit society. At the very
least, the existence of patents for a promising new technology often accelerates the process of
making that technology available to benefit the public by providing the necessary economic
incentives.
Recognition of the importance of the patent process in achieving successful application of new
inventions is the fundamental principle of the patent system and is a central feature of the laws
governing federally sponsored research in the U.S. For example, the Bayh-Dole act of the U.S.
Congress, which became effective in 1981, gives universities and small businesses the right to
claim ownership of patentable inventions that result from federally funded research. As a direct
result of the incentives created by this progressive legislation, there has been an explosive
growth in the patenting and licensing of university-based research results with several thousand
administrative support staff assisting these efforts across the United States. In turn, this has led to
the creation of numerous start-up companies, often involving university research staff. In an era
when funding from government sources is increasingly uncertain, the revenues returned to
universities through licensing of intellectual property contribute toward maintaining a strong and
healthy climate for applied, as well as for pure, research. Moreover, royalty revenues used to
support research generally allow greater discretion and flexibility compared to the more
commonly available directed research funds. Finally, given that a patent application can be
drafted and filed in the period between submission of a manuscript and the publication of the
paper, delay in the reporting of results may entirely be avoided. In summary, the need to publish,
and the need for patent protection (which will always remain a secondary objective in an
academic environment), are both crucial to society‘s interests and need not entail compromise.
Rebuttal
I find myself in agreement with several points discussed so eloquently by Dr. Sprawls in his
opening position. In particular, he concisely and accurately summarizes the importance, to
individual researchers and to society at large, of prompt presentation and publication of scientific
findings. Moreover, his statement, ‗‗The U.S. patent application process does not deter timely
publication of results if appropriate steps are taken for documenting research results‘‘, directly
supports a central theme of my position that delay in the publishing of results due to the drafting
and filing of a patent application may be entirely avoided.
However, the ‗‗conflict between publishing and concealing research findings,‘‘ mentioned in Dr.
Sprawls‘ opening position, is not something that normally enters into considerations of whether
to seek patent protection for new ideas before publishing articles about them, which is the
proposition to be addressed in this debate. The reason is that, in order to obtain protection for a
483
new idea through the filing of a patent, patent law requires the complete disclosure of the
concept—that is, nothing withheld from a patent application can be protected by a patent.
Therefore, ‗‗withholding valuable research findings from publication‘‘ would serve no purpose
vis-a`-vis obtaining patent protection since those findings would necessarily need to be disclosed
in the patent filing, which, if filed outside the United States or issued in the U.S. or elsewhere,
would become a public document. Of course, a researcher or his institution could decide to
protect an idea by choosing never to disclose it ~which would also necessitate never filing for
patent coverage!, thereby potentially creating a trade secret. In an academic environment,
however, obtaining trade secret protection would normally be inconsistent with the primary
objective of publication.
AGAINST THE PROPOSITION: Perry Sprawls, Ph.D.
Opening Statement
Virtually all mankind benefits today from the many advances in medicine and healthcare that
have occurred during the recent decades. This is especially true where physicists, other scientists,
and engineers have contributed to the development of imaging methods that lead to more
effective diagnosis and therapeutic procedures that reduce mortality and increase the quality of
life.
This has not come from a few researchers working in relative seclusion but from many in the
academic and industrial communities pursuing research and development projects.
Generally the objective of research is to extend the boundary of knowledge beyond what has
been established by other investigators. Without a comprehensive knowledge of prior research
results it is difficult to plan and execute effective research projects. Without this knowledge,
extensive research efforts are wasted on repeating investigations that have already been
conducted but not published by others. In many fields of research, scientists are quick to present
and publish results not only to enhance the global academic process but also to establish priority
and recognition for their research efforts. The additional value to the researcher who publishes
includes participation in scientific meetings, academic promotion, and access to funding.
Today, with much research directed to technology and process development, another issue arises
when the R and D results have financial value in the marketplace. This is the conflict between
publishing and concealing research results. While individuals and their organizations have a right
to financially benefit from their research efforts, this should not prevent timely publication. The
purpose of the patent process in our country is to protect the intellectual property of an individual
from unfair commercialization by others. It is not to be considered as a method of protecting
knowledge and research findings. The U.S. patent application process does not deter timely
publication of results if appropriate steps are taken for documenting research results.
There are many factors that should be considered by a researcher who is considering withholding
valuable research findings from publication:
• How will this information best serve humankind?
• Will the benefits of publication to me outweigh a remote possibility of financial gain
through the patent process?
• Is it even possible to get a patent on this?
• Does it really have a significant commercial value that should be protected by a patent?
484
• Am I or my organization willing to devote the resources (money, time, etc.) to the

patent process?
The conclusion is that research results should be published in a timely manner and not delayed
because of patent considerations.
Rebuttal
To publish or patent (and perish in the academic arena?), that is the question. Or is it the question
that should be debated here?
Dr. Antonuk and I both recognize the value of the patent process and also the opportunity for
academic recognition and the advancement of science and technology through the presentation
and publication of research findings.
In his opening statement he has clearly shown how patent protection contributes to the total
research and development process and can generate funding for on-going investigation. In many
cases this can be consistent with academic publication.
The real question to be considered is not so much publish or patent but how to publish and patent
so that neither is seriously compromised.
In response to Dr. Antonuk‘s thorough and compelling statement of support for the patent
process I remind us of the need for prompt publication. This not only serves the academic
aspirations of the individual scientist; it is one of the foundations of the total academic research
process.
485
10.18. A hospital-employed physicist working in radiology should

provide training to non-radiologists wishing to offer radiology
services
James A. Patton and Guy Simmons
Reproduced from Medical Physics, Vol. 26, No. 6, pp. 859-861, June 1999
OVERVIEW
Most hospital-based diagnostic medical physicists work principally within a radiology

department. In this setting, professional loyalties and personal friendships evolve among the
physicists, radiologists, and employees of the department. In most institutions, not all radiology
services are provided by radiologists or by the radiology department. In these institutions,
radiology services may be offered by cardiology, urology, obstetrics, cardiovascular surgery, and
other departments, frequently in direct competition with radiology. These departments need and
sometimes request the services of the medical physicist. The question for this issue of Point/
Counterpoint is whether physicists are ethically obligated to meet this need (even when their
services are not requested), or whether their professional relationships within radiology should
take precedence over the provision of services to competing specialties.
Arguing for the Proposition is James Patton. Dr. James Patton obtained his Ph.D. in medical
physics from Vanderbilt University in 1972 for work in single photon tomography and has been
on the faculty at Vanderbilt since that time. He currently holds the rank of Professor of
Radiology and Radiological Sciences and Professor of Physics. He is the chief nuclear medicine
physicist, has taught nuclear medicine physics and instrumentation to radiology and nuclear
medicine residents, nuclear medicine technologists, and cardiology fellows for 26 years, and has
served as program director for nuclear medicine technology for 20 years. Dr. Patton‘s research
interests include single photon tomography and applications of high energy imaging with dual
head scintillation cameras. He has co-authored five textbooks in nuclear medicine. He is
currently serving as President of the Southeastern Chapter of the Society of Nuclear Medicine.
Arguing against the Proposition is Guy Simmons. Guy Simmons, Ph.D., is a Professor at the
Department of Diagnostic Radiology, University of Kentucky Medical Center. He is also
President of Bluegrass Radiological Physics, Inc., a diagnostic and medical nuclear physics
consulting firm. Dr. Simmons received the Ph.D. in nuclear engineering from the University of
Cincinnati in 1972. He has served as Secretary and President of the American Association of
Physicists in Medicine. He is currently a Physics Trustee of the American Board of Radiology
and a member of the American College of Radiology Committee on Standards and Accreditation
of the Commission on Medical Physics.
For the proposition: James A. Patton
Opening Statement
486
Many institutions have made the decision to place radiologic technology into the hands of
nonradiologists for reasons that are often institution specific. Although these decisions are not
the subject of this discussion, it is clear that nonradiologists will continue to perform radiology
procedures.
Should radiological physicists train these specialists in basic radiation physics, instrumentation,
radiation safety, and quality assurance? The answer to this question is ‗‗yes.‘‘ The alternative is
for the specialists to train themselves or to attend concentrated courses outside of the institution.
Neither of these alternatives is acceptable. Each results in inadequate training which, although it
may meet the letter of credentialing requirements, certainly does not meet the spirit of the
credentialing process. Radiological physicists within institutions are better equipped to provide
the requisite training.
Every employee in a hospital, including the radiological physicist, has a responsibility to see that
patients receive the best care possible. Radiological physicists have a moral and ethical
obligation to see that procedures are optimized; equipment is monitored, maintained and utilized
properly; and radiation exposures are minimized, regardless of the user. Regulatory
organizations such as the Joint Commission on Accreditation of Healthcare Organizations
(JCAHO), state bureaus of radiological health, and the Nuclear Regulatory Commission place
stringent requirements on institutions regarding these topics. Radiological physicists have an
institutional responsibility to see that the requirements are met, again regardless of the user. As
hospital environments and patient care responsibilities continue to evolve, ongoing employment
for physicists in an institution may be contingent upon their willingness to train across
departmental lines.
This position is difficult for physicists who have worked side by side with radiologists for many
years. However, radiologists are being forced to form partnerships with other physicians in
orthopaedics, obstetrics and gynecology, nuclear cardiology, and interventional and vascular
radiology, to name a few examples. In our institution nuclear medicine physicians and
cardiologists share responsibilities in cardiac nuclear medicine (a hospital-based nuclear
medicine service) and in the nuclear cardiology laboratory (a cardiology outpatient facility).
Radiology-based physicists have responsibilities in both areas. Cardiology fellows attend
radiological physics classes with nuclear medicine and radiology residents and sit for an
examination at the completion of the course. They also rotate through the hospital-based cardiac
nuclear medicine laboratory and radiopharmacy. Although these agreements to share
responsibilities may not be ideal for radiologists, they appear to be necessary in order to survive.
Radiological physicists can play a significant role in ensuring the success of such partnerships.
Rebuttal
I agree with many of Dr. Simmons‘ statements. Radiologists are more qualified by training,
knowledge, and experience to perform radiologic procedures, and turf/economic factors are
driving the paradigm shift. I also agree that the allegiance of the radiological physicist is aligned
with radiology. However, the environment is changing and radiologists and physicists have no
control over the decisions that are being made. What is the future role of radiological physicists?
We can take the high road, as Dr. Simmons suggests, and simply say that we will not be a part of
a process in which we do not believe. On the other hand, we have an institutional responsibility
that probably will not permit us to make this decision. But more importantly, we have a
responsibility to patients to see that they receive the highest quality medical care involving
diagnostic and therapeutic radiologic procedures, regardless of the provider. Therefore I think the
487
decision is clear. Although we may not agree with the process, we have no choice but to assist in
the training of nonradiologists.
Against the proposition: Guy Simmons
Opening Statement
The pertinent questions surrounding the ethical dilemma of whether or not medical physicists
should aid and abet nonradiologists in their quest to perform procedures for which radiologists
are most qualified are the following. What is the driving force behind this encroachment? Is it
based on a firm conviction, supported by objective data, that the accuracy of diagnoses and thus
the quality of medical care will improve under this non-traditional approach? Or is the
encroachment a result of turf battles and therefore driven by economic factors. Most astute
observers would conclude the latter, which clarifies the dilemma considerably for physicists. If
we are to be true to our ethical standards we must support policies that require all medical
procedures to be performed by those who are best qualified and most competent in the respective
field. In the area of medical imaging, there is no argument that the competence lies with
radiologists, the only medical specialists that spend an entire five-year residency and devote their
professional careers solely to the diagnosis of disease using imaging procedures. The notion that
physicians in a different specialty can achieve a level of competence in clinical imaging
equivalent to that of radiologists through a process that amounts to on-the-job-training is
contrary to proven principles.
The time is long past in which one can become a competent medical physicist outside of a
structured, formal graduate program. The same is true of technical fields. How can we therefore
presume that diagnostic imaging procedures in the hands of physicians, whose formal training in
this highly complex discipline is minimal or nonexistent, will not degrade the quality of health
care? Most medical physicists are well aware of the scope and intensity of effort required to
achieve a foundation in imaging science necessary for a physician to master the specialty of
medical imaging. We know from experience that it cannot be accomplished in an anecdotal
fashion.
It has taken many years and much effort by physicians and physicists to bring radiology to the
state of prominence and respect it now occupies within the medical professions. For medical
physicists to engage in practices that undermine and destroy the fruits of this labor is a disservice
to our profession. Therefore the answer to the ethical dilemma is clear. As members of the
radiology team, we should not lend our support to a process that we believe will degrade, not
enhance, the quality of medical imaging services. The public we serve deserves better. Medical
physicists should continue to devote their physician training efforts to time-honored participation
in accredited residency programs for radiologists and continuing education for radiology
practitioners.
Rebuttal
Dr. Patton makes several cogent points in his customarily elegant fashion. I will quote three
passages from his discourse to support my counterpoint argument. Dr. Patton, whom I know to
be a man of the highest ethical and moral standards, makes the following statement
commensurate with his own principles; ‗‗Every employee in a hospital, including radiological
physicists, has a responsibility to see that patients receive the best care possible. Radiological
physicists have a moral and ethical obligation to see that procedures are optimized . . . .‘‘ What
488
better way to achieve this worthy goal than to insist that diagnostic imaging procedures be
performed by the physicians best trained to do them; viz. radiologists? In his last paragraph Dr.
Patton expresses a highly pertinent truism, ‗‗. . . radiologists are being forced to form
partnerships with other physicians . . . .‘‘ The key word here is ‗‗forced,‘‘ implying that the
migration of medical imaging to other specialities is driven by market conditions, not because it
enhances the quality of medical care. Finally, he admits, ‗‗Although these agreements to share
responsibilities may not be ideal for radiologists, they appear to be necessary in order to
survive.‘‘ I agree that the survival of radiology as a medical specialty is essential to the delivery
of quality medical care. But capitulating to the forces that threaten radiology is not the way to
ensure survival. As radiological physicists we are better advised to oppose those who seek to
fragment medical imaging and support efforts to keep it where it belongs—in the hands of
radiologists.
489
10.19. The Joint Commission for the Accreditation of Healthcare

Organizations should require that all physicians who perform
fluoroscopy be credentialed in radiation management by their
healthcare facility
Benjamin R. Archer and David S. Gooden
Reproduced from Medical Physics, Vol. 27, Issue 1, pp. 1-3, January 2000
OVERVIEW
Patient exposures in fluoroscopy for specific examinations vary widely among physicians. Some
physicians cause high exposures because they have a ‗‗heavy foot‘‘ or because they take an
inordinate time to perform fluoroscopic examinations or fluoroscopy-guided procedures.
Exposures are lower with other physicians, perhaps because they are more conscious of the
exposure of patients to radiation, and also because they may be better trained in the efficient use
of radiation. There is some suspicion, not well documented, that nonradiologists tend to produce
higher patient exposures than radiologists. An initiative proposed from time to time to remedy
the cause of high fluoroscopic exposures is to require on-the-job training of physicians before
they are credentialed to perform fluoroscopy in the hospital. Some physicists support this
initiative, while others believe it will not be productive and require too much time and effort
from medical physicists. This is the subject for this issue of Point/Counterpoint.
Arguing for the Proposition is Benjamin R. Archer, Ph.D. Dr. Archer is currently an Associate
Professor of Radiological Sciences at Baylor College of Medicine where he has been employed
since 1984. He received his doctorate from the University of Texas Graduate School of
Biomedical Sciences/M. D. Anderson. He has served on numerous AAPM committees and task
groups. Currently he co-chairs the NCRP 49 Rewrite Committee and serves as Treasurer of the
American Board of Medical Physics. He is also active in the American College of Medical
Physics and the American College of Radiology.
Arguing against the Proposition is David S. Gooden, Ph.D., J.D. David S. Gooden is the Director
of Biomedical Physics at Saint Francis Hospital in Tulsa. For almost 30 years he has served as
radiation safety officer and radiological physicist for diagnostic x-ray, radiation therapy and
nuclear medicine. Dr. Gooden has provided radiation safety consultation in many areas,
including health care, veterinary medicine, nuclear reactors, electric utilities, universities,
industrial radiography, waste management, scrap metal salvage, foundries, and oil and gas
production. Dr. Gooden is certified by ABHP (health physics), ABR (radiological physics), and
ABMP (radiation oncology physics).
FOR THE PROPOSITION: Benjamin R. Archer, Ph.D.

Opening Statement
Massage therapists, beauticians and realtors must be licensed in order to practice in most states.
However, little or no training is required for most physicians who daily expose their patients,
staff and themselves to ionizing radiation! Fluoroscopic radiation is known to have caused
490
cancer and severe injuries in a small but growing population of physicians and patients. More
than fifty radiation burns, including a very recent one that produced a deep necrotic ulcer and
ultimately rendered the humerus of a patient visible (Fig. 1), have been reported in the United
States. A 1994 advisory notice issued by the U.S. Food and Drug Administration warned
facilities of the potential dangers of invasive fluoroscopic procedures. The notice recommends
that facilities ‗‗assure appropriate credentials and training of physicians performing
fluoroscopy.‘‘
FIG. 1. Radiation burn resulting from debridement following radiofrequency

cardiac catheter ablation procedure. Injuries of this magnitude are very rare.
This case will be discussed fully in an article in the first edition of the
Journal of the American College of Medical Physics in 2000. (Photo courtesy
of L. K. Wagner)
Medical staff credentialing is the verification and assessment of practitioner qualifications and
the granting and delineation of clinical privileges. The credentialing process was instituted to
protect patients from unethical or untrained practitioners by recognizing the competence of a
professional.
Recognition of competence may be through certification, licensure, registration or a combination

of the three. It nearly always involves the completion of an accredited or approved educational
program. The credentialing of a medical staff member implies verification of competence. There
is a wide distinction, however, between the credentialing of a cardiologist, radiologist, pain
interventionalist, or other fluoroscopist as technically competent to perform a procedure versus
the credentialing of the same physician as competent to safely use fluoroscopic radiation.
Credentialing these physicians as technically competent gives no assurances that they have
received training in the safe uses of fluoroscopy. I submit that the current ‗‗rubber stamp‘‘
granting of privileges to untrained practitioners is both illogical and unethical. The fact is that
these individuals can and are injuring patients and even one preventable radiation injury caused
by ignorance and indifference is absolutely unacceptable!
Furthermore, as the pre-MQSA experience has shown us, unless credentialing is adequately
enforced, the results will be less than satisfactory. Enforcement should logically come from the
JCAHO. Dr. J. B. Spies clearly elucidated the rationale at the 1992 ACR/FDA Workshop on
Fluoroscopy: ‗‗Since all health facilities require accreditation by the JCAHO in order to receive
Medicare and Medicaid funds, adding a requirement for fluoroscopy certification would in
essence be self-enforcing by the individual facilities.‘‘ Thus, if the JCAHO were to require
491
appropriate credentials for fluoroscopists in radiation management, then each healthcare facility
would have to see that its staff complied or else be faced with the unseemly prospect of forfeiting
revenue. Concern about malpractice litigation should also provide convincing motivation for
facilities to proactively seek credentialing and to welcome the JCAHO influence and audit. But
the major motivation should result from plain common sense and the innate ethical realization
that we should always do the ‗‗right thing‘‘ for our patients.
Rebuttal
Physicists can write prescriptions for patients! Surgeons can interpret mammograms! Red Goose
shoes will rule the world (I wore them too)! David, these statements are as nonsequitur as your
contention that it is wasteful to spend money on credentialing fluoroscopists. Radiation injuries
caused by physicians are not a Chicken Little story. The cost of education pales against the
human suffering and expense of malpractice lawsuits. We are presently aware of at least five
such trials that are in progress. There have been and will be more.
Fluoroscopic equipment has become more powerful and complex. Longer and more difficult
procedures have become commonplace. To illustrate just how essential it is to credential
fluoroscopists in radiation management, Dr. Lou Wagner and I just completed a study of
radiation doses from two modern special procedure fluoroscopic units (JVIR, in press). We
found that a prolonged procedure on a large patient performed with less than ideal radiation
management techniques resulted in a dose of 8.8 Gy more than the same exam performed with
ideal techniques! In other words, physicians who are ignorant of practical physics concepts and
instrumentation can unwittingly produce erythema and dermal atrophy in their patients. The
boogey man here is apathy.
Let‘s be clear on what we mean by credentialing because I believe we are not that far apart. As
you suggest, ‗‗better training in medical schools, residencies and good institutional radiation
safety programs can reduce fluoroscopic injury to near zero.‘‘ Completion of an approved
educational program (with appropriate testing) provides the evidence needed by the facility to
approve the practitioner‘s qualifications. A physician can be credentialed simply by completing
an appropriate course(s) in radiation management. The JCAHO audit would insure that program
goals are being achieved.
Again, even one radiation injury caused by lack of education is unacceptable. An attitude change
is required. It should become our quest as radiation experts to assure that the words reportedly
spoken by a cardiologist in a recent trial regarding a radiation burn on his patient are never again
heard: ‗‗I only wish I had known that I could have done that to my patient; I thought a burn like
that could only happen in radiation therapy.‘‘
AGAINST THE PROPOSITION: David S. Gooden, Ph.D., J.D.

Opening Statement
The earth is flat! The sun and stars rotate around the earth! Lead can be changed into gold!
Radiation is so dangerous that it requires extraordinary controls and regulations! These beliefs
are based on scientific truths that existed at the time of the belief. Three beliefs are abandoned
and now seem foolish. The fear of radiation, however, remains deeply ingrained in us. This fear
is traceable to a foundationless
492
linear-no-threshold model of radiation injury promoted by activists who in the 1950s sought to
stop testing of nuclear weapons. Requiring more extraordinary controls such as special
credentials for physicians using fluoroscopy propagates this unreasonable fear of radiation.
My position on special credentials comes from a lifetime association with radiation and the study
of radiation injury. As a small boy (late ‘40s and early ‘50s), our family doctor would invite my
mother and sister to observe my body with an upright fluoroscopy device (estimated 25–50
R/min). I remember my mother leaving me at the Red Goose Shoe Store in Atlanta while she
shopped. My sister and I played with the x-ray shoe-fitting machine for hours (estimated 3000
R/hr). While at Emory University, I worked with an early prototype Co-60 device and performed
dosimetry on a kilocurie Cs137 irradiation facility. At the University of Missouri Research
Reactor, I had safety responsibilities for kilogram quantities of 98% enriched U-235 and
megacuries of spent fuel. For 35 years, I have helped both medical and nonmedical industries use
radiation in a safe, productive manner. My study and experience taught me that large acute doses
of radiation could cause severe injury and death. Also, there is evidence that radiation is a weak
carcinogen for moderate exposures (10s to 100s cGy) at high dose rates.
Radiation is dangerous and I do not want to return to the lax practices of the 40s and 50s.
However, radiation is not extraordinarily dangerous. Rather, it is dangerous like gasoline, fire,
explosives, electricity, hammers and saws, syringe needles, and all other tools. It is dangerous
like water, oxygen, vitamins, drugs, chemicals, food, sunshine, sleep, and all other things that
cause harm if exposures are too great.
Radiation, like all things that serve humankind, must be used with caution. Although new life
saving interventions requiring long fluoroscopy times have caused a few dozen acute injuries
during the past few years, I believe that radiation medicine remains the safest of the major
branches of health care. Other branches of medicine kill and maim thousands each year.
Reasonable control and regulation of radiation is appropriate, but extraordinary controls are not
required. Better training in medical schools and residencies, and good institutional radiation
safety programs, can reduce fluoroscopy injury to near zero. To spend money on special
credentials for fluoroscopists is wasteful and sends the erroneous message that radiation is
extraordinarily DANGEROUS. We can serve our patients best by using our limited resources
where the greatest benefit can be realized rather than by throwing more money at the radiation
boogey man.
Rebuttal
My colleague, Ben Archer, makes a good argument in favor of certification of all physicians who
perform fluoroscopy. Dr. Archer and I agree that acute fluoroscopy injury can be, and should be,
reduced to (or near to) zero. Our beliefs in the root causes of injuries (and, thus, the cure),
however, differ. Ben appears to believe that the injuries are caused by untrained and/or unethical
physicians. I believe otherwise. The injuries that I know of in Oklahoma have not occurred at the
hands of untrained, unethical physicians. In fact, they were caused by radiologists and
cardiologists well-trained in interventions requiring fluoroscopy. Far from being the result of
unethical practice, they resulted from interventions without which the patients would likely have
died. I believe most of the recent fluoroscopy injuries have resulted from the ethical practice of
trained physicians.
493
The root causes of the recent fluoroscopy injuries are 1) inadequate early supervision of residents
in training programs, 2) the learning curve associated with new life saving interventions that
require long fluoroscopy times, and 3) the intense concentration on the procedure to the
exclusion of noting important fluoroscopy parameters. Important parameters include: 1) the
distance between the patient and the x ray tube, 2) the x ray output mode (normal or
supercharged), 3) time (use freeze frames and/or light foot), and 4) the movement of the x-ray
entry point to different locations on the patient. The entire team assisting in the intervention must
be encouraged to stay cognizant of fluoroscopy parameters to maximize benefit and reduce risk.
Since untrained and unethical physicians are not the root cause of fluoroscopy injuries, safety
will not be improved with requiring more credentials. Requiring extra credentials will, however,
promote the unreasonable belief that radiation is extraordinarily dangerous and, thus, distract
from the medical use of this wonderful tool. Root causes of fluoroscopy injuries should be
addressed in training programs and institutional safety programs. The curriculums of medical
schools and residencies should teach that radiation is like a surgeon‘s scalpel. It is itself neither
good nor evil, but instead a tool that can be used to great benefit if used with reasonable (not
extraordinary) caution. Curricula should include early training in reasonable caution. Directors of
institutional radiation safety programs should promote a partnership with physicians in the safe
and beneficial use of medical radiation for diagnoses and interventions.
494
10.20. The Ph.D. degree is a handicap in the job market for clinical
medical physicists
J. Daniel Bourland and David S. Marsden
Reproduced from Medical Physics, Vol. 27, No. 12, pp. 2641-2643, December 2000
OVERVIEW
The Ph.D. degree signifies that the individual has the ability to pursue an original investigation
designed to generate new knowledge in a particular field. It is the appropriate degree for the
person interested in an academic career in the discipline. For skilled application of existing
knowledge in a clinical arena, however, some may view the Ph.D. as undesirable because the
degree connotes greater interest (and possibly knowledge) in research than in application.
However, an individual with the Ph.D. degree has met higher academic standards, and may be
viewed more as an equal with other professional in a clinical setting. The relative merits of the
Ph.D. degree compared with a less research-oriented degree (e.g., the MS degree) in seeking a
nonacademic position in medical physics is the subject of this Point/Counterpoint.
Arguing for the Proposition is J. Daniel Bourland, Ph.D. Dr. Bourland is Assistant Professor and
Head, Physics Section, in the Department of Radiation Oncology, Wake Forest University,
Winston-Salem, NC. He is Chair of the AAPM‘s Electronic Media Coordinating Committee and
the African Affairs Sub-Committee. He performs clinical service, research, and education in the
areas of 3D treatment planning, radiosurgery, and the uses of images in treatment planning.
Arguing against the Proposition is David S. Marsden, Ph.D. After thirty years as the Radiation
Safety Officer and Director of both Medical Physics and the Radioimmunoassay Laboratory at
St. Luke‘s Roosevelt Hospital in New York City, Dr. Marsden recently retired to form a family
operated consulting firm, Marsden Medical Physics Associates LLC. Dr. Marsden has been a
physics advisor and oral and written board examination committee member for the ABR. He was
the Chairman of the AAPM Training of Radiologists Committee. He helped create the Clinical
Ligand Assay Society and served this organization in every elected position. He was recently
awarded fellowship in the ACR.
FOR THE PROPOSITION: J. Daniel Bourland, Ph.D.
Opening Statement
The Ph.D. degree is a handicap in the job market for clinical medical physicists because it
reflects the character of the individual who has taken the time to earn the degree. This character,
in general, makes some Ph.D.s ill-suited for clinical medical physics work, and this trend is
known by administrators and other physicists. Thus, a certain stigma is associated with the Ph.D.
degree that is a handicap for recruitment to clinical medical physics positions. The Ph.D. medical
physicist is an individual who has interrogated a research topic at great depth. By nature, the
Ph.D. shows that one is curious about his environment and interested in investigation through
theoretical and experimental methods. Thus, Ph.D. medical physicists may not be satisfied with
the completion of routine quality assurance work—a daily supply of work is not the ticket to
495
contentment. This inherent discontentment makes the Ph.D. one who is difficult to please
because challenge and variety are either not present, or possibly not accommodated in a clinical
position. While critical thinking is an attribute for medical physicists in the clinic, the tendency
for the Ph.D. to ponder a situation can result in over analysis and little solution. Also, many
Ph.D.s have concentrated on their research during their educational years, and have little clinical
or practical experience. Such individuals do not belong in the clinic for reasons of efficiency and
safety for themselves, co-workers, and most importantly, patients. Another difficulty with the
Ph.D. is the salary expectation, which runs 10 to 20% higher than for MS-level positions. So the
Ph.D. will cost more than the MS in the same clinical position.
Some Ph.D. medical physicists serve whole-heartedly in clinical positions. In general, however,
administrators and co-workers know that the Ph.D. medical physicist will (1) eventually get
bored with routine clinical work, (2) be interested in detailed research projects with doubtful
clinical application, (3) ask to do some other kind of work when clinical service is required
(redefine the position), (4) ask for higher compensation, and (5) be difficult to please in other
facets of employment. These common characteristics are a handicap to the Ph.D. seeking a
clinical medical physics position, and provide a clear distinction between Ph.D. and MS level
positions.
Rebuttal
Dr. Marsden‘s response to the Ph.D. being a handicap lists advantages of the Ph.D. degree.
These advantages may be valuable for the Ph.D. medical physicist, but they do not apply to the
routine clinical environment. Higher earning potential, credibility, and a balanced working
environment (‗‗Doctor‘‘) may make one‘s job professionally more satisfying, but distract from
the duty at hand, which is clinical medical physics. While academic institutions need medical
physicists at the Ph.D. level, community settings need people who will do a variety of routine
tasks. Note that ‗‗routine‘‘ does not mean mundane, though this is sometimes the case. Even
academic institutions hire MS-level medical physicists in their clinics to provide most of the
clinical service and to free up Ph.D. time for teaching and research. As discussed by Dr.
Marsden, clinics hiring MS medical physicists may pay less than a Ph.D. However, given today‘s
competitive job market, this differential is becoming less and less the case. Certainly, board
certification is a primary salary differential and reflects an important point—the MS degree for
medical physics is a perfectly valid level of education. The AAPM offers educational fellowships
to encourage clinical medical physics training at the MS level, and the MS is acceptable for
gaining entrance to board certification exams. MS medical physicists can enhance their careers
by serving in administrative positions, directing allied health training programs, or serving as the
sole institutional physicist. Thus, a pathway of service is an alternate to the academic track.
For these reasons and those stated earlier, clinical medical physics positions are well covered by
MS medical physicists, and the Ph.D. medical physicist is best matched to the academic
environment. I propose that the best clinical medical physicist is one who has either the MS or
Ph.D. degree, has strong analytical skills, is dedicated to clinical service, shows his expertise by
board certification, and has strong computer skills since our field is becoming heavily
computerized. Credibility will be earned based on one‘s skills and wisdom in responding to
clinical situations, not on the level of degree.
AGAINST THE PROPOSITION: David S. Marsden, Ph.D.

496
Opening Statement
Although a Clinical Medical Physicist can work independently and become board certified with
only a master‘s degree, a doctoral degree is an advantage in the job market. Even though
individuals with a master‘s degree may be hired more readily in some environments, their ability
to attain higher levels of achievement are compromised. To be on the medical staff at a hospital,
a Ph.D. is required. To achieve Full Professor or Department Chair status in an academic setting,
a Ph.D. is often a prerequisite. The doctoral degree enables physicists to choose from a broader
range of employment opportunities. Maybe that‘s why in 1999, masters level Medical Physicists
were 11/2 X more likely to change jobs when compared with doctoral level Medical Physicists. 1
The salaries of Clinical Medical Physicists with a doctoral degree are higher than the salaries of
Clinical Medical Physicists with a master‘s degree. In 1999, physicists with a doctoral degree
and board certification had an average income of $116,500, whereas, physicists with a masters‘
degree and board certification had an average income of $100,900. Physicists with a doctoral
degree and no board certification had an average income of $93,800; whereas, physicists with a
masters‘ degree and no board certification had an average income of $81,300 (AAPM Survey,
1999). Fifty-six percent of AAPM members have a Ph.D. degree and thus, benefit from the
possibility of higher earning potential.
In addition, many Clinical Medical Physicists work in the medical community, which highly
regards the title ‗‗Doctor.‘‘ Being able to use the title ‗‗Doctor‘‘ in a medical setting creates a
situation in which Physicists are seen as credible competent professionals (which may or may not
be true). The title affords physicists the opportunity to work on an even playing field. Most
people do not understand the title Medical Physicist and when we talk in terms of working with
radiographic or mammographic units they confuse us with service technicians. Most patients
don‘t realize that Physicists are part of their radiation treatment team. Like it or not, the use of
the title ‗‗Doctor‘‘ connotes importance and causes people to take our profession and us more
seriously.
To complete a doctoral program, an individual must engage in several years of research. The
application of the scientific method to a research problem teaches students to think critically and
logically. Thus, students gain more than just rote on-the-job training. They are taught to analyze
the methods they are using and evaluate all possible options, which makes them more
marketable. An employer can expect an individual with a doctoral degree to possess versatile
problem solving skills as well as specific technical information.
The additional education required to obtain a doctorate affords several benefits for later
employment. The ability for enhanced career attainment, higher earning potential, and
professional credibility, creates a more balanced working environment.
Rebuttal
Research done while completing a doctoral degree is an exercise in the use of the scientific
method and analytical skills. Just because an individual possesses the skills to complete a
research project does not necessarily indicate he or she has a similar ‗‗character‘‘ or even a
desire to work as a researcher. Students have a myriad of reasons for attending graduate school
and developing different proficiencies in research and clinical work depending on their interests,
program attended, and completion of clinical internships, etc. Indeed, there may be some medical
physicists who become ‗‗bored with routine clinical work,‘‘ or attempt to ‗‗redefine the
497
position.‘‘ However, these tendencies are probably based more on an individual‘s need for
challenges or on the nature of the work environment, rather than on the individual‘s educational
background.
Obviously, physicists without practical experience should not be working in clinical positions.
Whether or not an individual is competent to function in an applied physics position is based
more on professional experience than on academic degree. Master‘s degree programs are not set
up to offer more clinical experience than doctoral programs.
My colleague asserts that ‗‗the character of the individual who has taken the time to earn the
degree (Ph.D.)...makes some Ph.D.s ill suited for clinical medical physics work, and this trend is
known by administrators and other physicists. ... .‘‘ I am unaware of any data (statistics, AAPM
publications) that support this assertion. It seems to be a generalization based on opinion rather
than fact. If it were true, then very few Ph.D.s would be employed, which is obviously not the
case.
REFERENCE
1. American Association of Physicists in Medicine, ‗‗Professional Information Survey

Report, Calendar Year 1999,‘‘ 2000.
498
10.21. Medical physicists need professional malpractice insurance

Michael Davis and Jeffrey Masten
OVERVIEW
Many medical physicists work full-time as independent contractors, sometimes as a member of a

physics consulting firm. Other physicists work full-time as an institutional employee, but provide
consultative physics services part-time as a source of supplemental income. For all of these
individuals, the cost and breadth of coverage of malpractice insurance is a key concern. Some
physicists believe that coverage is essential regardless of cost. Others choose to go without
coverage because of the cost and the rationale that malpractice insurance enhances the risk of
lawsuit. The controversy over malpractice coverage is explored in this point/counterpoint.
Arguing for the proposition is Michael Davis, M.S., J.D. Mr. Davis received his Masters degree
in Radiological Medical Physics from the University of Kentucky. He recently left his position as
Director of Radiation Oncology for the Greenville Hospital System in South Carolina to join
Varian Medical Systems. He also worked at the M.D. Anderson Cancer Center in Houston for
eight years. During that time he obtained a Doctor of Jurisprudence degree from the South Texas
College of Law. Mr. Davis is certified in Therapeutic Radiological Physics by the ABR. He
currently chairs the Legal Affairs and Risk Management subcommittee of the AAPM.
Arguing against the Proposition is Jeffrey Masten, M.A., J.D. Mr. Masten is a Staff Physicist at
the Medical X-Ray Center, P.C. in Sioux Falls, South Dakota. He earned his M.A. in Physics at
Columbia University, NYC, in 1972, his J.D. at the University of South Dakota in 1976, and is an
M.S. candidate in Medical Physics at the University of Colorado, Denver. He has been a trial
lawyer for 25 years with a particular interest in the use of scientific evidence in litigation.
FOR THE PROPOSITION: Michael Davis, M.S., J.D.
Opening Statement
We are potentially victims of our own success. The technologies used in radiology and radiation
oncology clinics have advanced dramatically during the last few years. This rate has increased
the need for physics services. This is one cause of the recent spiral in salaries of medical
physicists. Utilizing new technologies and making more money can be exciting and rewarding.
Unfortunately these very happenings make us tempting targets.
All of us take pride when our work positively touches the lives of people. Countless patients owe
at least part of the detection or treatment of their illnesses to our professional efforts. But
occasionally a failure occurs. In some of these failures, a patient or family member may decide
that the efforts of the healthcare team fell short. And he or she may seek monetary compensation.
In a perfect world there would never be a treatment error. Unfortunately, this is not the world in
which we work. Sometimes an error that affects a treatment outcome gets past us. And
499
sometimes someone sues because of the outcome. When this happens our fate is left in the hands
of an attorney and a judge or jury of peers.
Anyone who has ever been involved in a lawsuit knows that it generates enormous stress. This
stress is greatest when one is on the wrong end of the suit. A lawsuit may drag on for years. An
adverse decision in a malpractice case can destroy much of what has taken years to build. Even if
one is ultimately victorious, the legal costs can be staggering.
Obtaining malpractice insurance is one cost of doing business today. Our dependence on
computer software, elaborate technology, and other professionals on the team put us at risk.
Participation in increasingly complex treatment methods, such as IMRT, has compounded this
risk.
Because we are not the people taking direct responsibility for patient care—i.e., we are not
physicians—the malpractice insurance costs for physicists tend to be relatively reasonable. In
fact, as a percentage of what one stands to lose without coverage, it is trivial. Even if the cost of
professional malpractice insurance was much higher, cost trimming or "going bare" would not be
wise.
When we reach a point where the legal system, every one of our colleagues, and we ourselves as
physicists are perfect, then we can dispense with professional malpractice insurance. That time
has not yet arrived.
Rebuttal
Mr. Masten believes that liability is the employer's problem. Ideally this is true. But there are
several factors that make relying upon the good graces of employers a mistake. We are in a
dynamic healthcare environment. Who has responsibility for our malpractice defense if our
employer closes its doors or merges with another organization? Often a lawsuit is not initiated
until years after the initial causal event, because the plaintiff may not experience symptoms for
some time. What if we have changed employers during that period? And how many of us even
know if the group policy provided by our employer is adequate?
As Mr. Masten states, moonlighting or solo practices are clearly exceptions to relying upon an
employer-sponsored master policy. But what about the gray areas that arise even if we are not
involved in these activities? We remain qualified medical physicists even outside of work. What
happens if we offer advice to friends or strangers at a social event? Would our employer have any
responsibility for defending us if those to whom we gave information decide that we have caused
damage through poor professional advice?
Mr. Masten worries about finger pointing and hard feelings on the part of the employer if we
have a separate malpractice insurance carrier and decide to give notice to that carrier. But that is
a small price to pay for having our own attorney who has our interests at heart, rather than simply
relying on an attorney assigned to us by the employer. Counsel for the employer will try to
protect the overall interests of the employer, and may have to balance those interests against the
individual needs of the employee. By contrast, the physicist's own attorney would be focused
solely on what is in his or her best interest.
500
Physicists are familiar with the concept of weighing risk versus benefit. I believe that the
potential threats created by not having professional malpractice insurance far outweigh the
financial costs of obtaining it. As Mr. Masten said, better safe than sorry is an admirable attitude.
AGAINST THE PROPOSITION: Jeffrey Masten, M.A., J.D.
Opening Statement
My comments are directed toward physicists employed by an entity (e.g., a hospital, clinic or
group practice) that maintains a master policy providing malpractice coverage for its employees.
Moonlighting and solo practice are excluded. Should a physicist spend the money to have an
independent policy separate and apart from the employer's?
One frequently heard argument against malpractice insurance is that the coverage makes the
insured a "target" for lawsuits. As we shall see, that is a poor justification for not purchasing
malpractice coverage because anyone participating in the treatment of a patient that results in a
malpractice claim is likely to be named in the lawsuit. Plaintiff's counsel will not pick and choose
defendants based on insurance coverage, especially when failure to identify an indispensable
party to the suit may result in dismissal of the claim altogether.
The dominant factor in a decision to secure individual coverage must be consideration of the
likelihood that the employee, but not the employer, will be sued. The employee must further
assess his or her ability to pass the defense onto the employer should a lawsuit occur. This is the
point at which the current drive toward certification and recognition of the physicist as a
professional who exercises independent judgment becomes important.
If the employer can tell an employee how to do his job, liability is the employer's problem. That
is the gist of the master-servant relationship. If, however, the employee exercises independent
judgment, the employee may be considered an independent contractor. In this case the liability
may reside exclusively with the employee. So long as employment falls within the master-servant
doctrine, additional malpractice coverage is superfluous. There is no reported case in which an
employer has argued successfully that it could not control a physicist's actions, and therefore that
the master-servant relationship did not exist.
There may be additional fallout from securing a separate policy. A professional malpractice
policy requires that the insured give notice of any event that may result in a claim. If you decide
to be especially conservative and give notice to your carrier, but your employer has a different
view, there may be some hard feelings even if the court decides (as it most likely would) that the
employer's carrier cannot decline coverage for failure to give notice. With multiple carriers
involved, settlement negotiations can easily degenerate into finger pointing and recrimination,
especially if the right personalities and large potential damages are involved.
Better safe than sorry is an admirable attitude. But before a physicist buys a separate policy, it
may be best to talk it through with an experienced attorney.
Rebuttal
The most attractive argument in favor of obtaining professional liability coverage is based on a
risk/benefit analysis, i.e., the amount you could lose in court pales in comparison to the cost of
501
the insurance. While I agree with Mr. Davis that every physicist should be covered by some form
of liability insurance, I do not agree that this necessarily requires an individual policy in all
situations.
Physicists in solo practice, working in a group that does not have a master policy, or
moonlighting as an employee of some other organization, definitely need individual coverage.
For the rest of us, the decision depends in part on an analysis of our employment contract, the
laws of the jurisdiction where employment takes place, and a realization that double coverage
doesn't always mean double protection. An office visit with a competent local attorney is the
place to start.
The bottom line is "Who should pay for the peace of mind which might be afforded by a separate
professional liability policy?" I don't see why I should dig into my bank account for protection
from my employer. Mr. Davis is correct that professional liability coverage is part of the cost of
doing business today. However, if 100% of my effort is devoted to my employer's business, then
it's not unreasonable to expect the employer to pay 100% of the cost of my liability coverage.
502
10.22. Medical physicists should not publicly endorse commercial

products
Michael G. Davis and D. Jay Freedman
OVERVIEW
From time to time, a medical physicist may have an opportunity to evaluate a technology early in
its development or deployment. If the evaluation is positive, the company responsible for the
technology may ask the physicist to endorse it, or at least be identified as an evaluator. The
reward may be a financial contribution to the physicist or the institution, or may simply be
enhanced personal or institutional recognition. Some physicists see nothing wrong in an
endorsement, whereas others believe it undermines the physicist's integrity and credibility. These
points of view are discussed in this Point/Counterpoint.
Arguing for the proposition is Michael Davis, M.S., J.D. Mr. Davis received his Masters degree
in Radiological Medical Physics from the University of Kentucky. Currently he is the Director of
Radiation Oncology for the Greenville Hospital System in Greenville, South Carolina. He
worked at the M.D. Anderson Cancer Center in Houston for eight years prior to moving to
Greenville. During his time in Houston he obtained a Doctor of Jurisprudence degree from the
South Texas College of Law. Mr. Davis is certified in Therapeutic Radiological Physics by the
ABR. He currently chairs the Legal Affairs and Risk Management subcommittee of the AAPM
and is also a member of the ethics committee.
Arguing against the proposition is D. Jay Freedman, M.S. Mr. Freedman received his M.S. in
Physics from the University of Alabama in Birmingham in 1979. In 1987, he spent a year as
guest physicist at Rambam Medical Center in Haifa, Israel. The following year he enrolled at
Hebrew Union College in Jerusalem while continuing to lecture in Medical Physics at Haddassah
Medical Center. He received Rabbinic Ordination in 1994 at the Hebrew Union College in
Cincinnati. Since then he has performed various pastoral duties as well as working in Medical
Physics. Freedman was appointed chairman of the AAPM Ethics Committee in January 2001. He
has been a senior physicist at Hartford Hospital since March 1999.
FOR THE PROPOSITION: Michael G. Davis, M.S., J.D.
Opening Statement
No one really believes that Tiger Woods is a champion just because he uses a particular brand of
golf ball. We know he endorses the manufacturer of that golf ball at least partially because of
lucrative offers of money to do so. Tiger Woods is an entertainer and endorsements are part of
what an entertainer does for a living. Medical physicists are not entertainers. Our professional
goals and obligations are very different. We claim that our evaluations of equipment and vendors
are based upon objective scientific reasoning. But can these opinions be truly objective, or
503
viewed as truly objective, when companies are offering us perks in exchange for public
endorsements?
When we publicly endorse a product or company, we cross a line. We are no longer supplying
information to colleagues who seek us out for an opinion through a phone call or professional
publications. Instead, we are voicing an opinion that most of our colleagues will hear. Often
behind this endorsement is free or bargain-priced equipment, money for research, lucrative
consulting contracts, expensive dinners, or invitations to speak at vendor functions. Nothing may
be wrong with receiving any of these things per se. But when these benefits are perceived as
payments for product endorsement, our individual integrity, and our credibility as a profession
suffer.
The chief victims of paid endorsements may be the less experienced medical physicists among
us. They may not have gained all of the background and tools needed to make objective choices
regarding vendors. It is one thing to inform them by using peer-reviewed equipment or software
comparisons within the pages of Medical Physics. It is quite another thing to expose them to
quotes of a well-respected senior physicist emblazoned upon a vendor's booth at an AAPM
meeting.
Medical physicists have a responsibility to keep vendors honest. We must make sure that a
vendor provides the product and support promised at the time of purchase. We must also make
sure that we are honest with our colleagues regarding vendor choices before them. This is how
we improve products and service, which lead to improved patient care. Credibility is the key to
all of this. Our job is made more difficult when our credibility as physicists is compromised by
product endorsements from our colleagues.
Most of us would be flattered to see our photo on an advertisement. Most of us would love to
have our opinions widely quoted for our friends and colleagues to read. However, we chose a
different path when we decided to become medical physicists. I believe that this choice requires
that we not allow ourselves to become marketing tools. After all, we are not entertainers.
Rebuttal
I agree with Mr. Freedman that medical physicists should receive fair remuneration for their
work, but he and I philosophically part company after that. Professionals must never forget for
whom they work. The attorney-client privilege exists in law because the client is paying for the
exclusive use of the professional services of the attorney in a particular matter. A lawyer who
forgets this and shares confidential information with opposing council, or anyone else for that
matter, has breached an ethical duty because doing so may weaken the client's case.
However Mr. Freedman indicates that we can serve two potentially opposing interests
simultaneously if we are careful so as not to create a true conflict of interest. These two opposing
interests are the employers of our medical physics services and the commercial vendors selling
products to those employers. I contend that the act of trying to serve two opposing masters is by
its very nature a true conflict of interest. There is no fine line to be walked. There is however a
clearly marked boundary to be avoided.
Mr. Freedman also says that we as medical physicists have no Consumer's Union to turn to for
advice in product evaluation. I would contend that collectively we are in fact a Consumer's Union
for medical physics and radiation therapy products. We test and evaluate equipment constantly
504
and most of us are very willing to express our opinions about the strengths and weakness of a
product to whomever asks.
Finally, I disagree with the premise that there is an obligation to provide public endorsements of
products. By its very nature an endorsement avoids mentioning flaws in a product or service
while it enhances the good. Part of the reputation we build professionally with colleagues is in
our willingness to mention these flaws in our exchanges of information with them. To leave out
the bad information damages this reputation. And as Joseph Hall said "A reputation once broken
may possibly be repaired, but the world will always keep their eyes on the spot where the crack
was."
AGAINST THE PROPOSITION: D. Jay Freedman, M.S.
Opening Statement
In the 1970s there was a television program called Mary Hartman, Mary Hartman. During one of
the first episodes, the title character said in earnest, "Everything I know I learned from
commercials."
I hope and expect that we Medical Physicists are more discerning than Mary Hartman. However,
there is much to be said for the educational value of advertisements we receive in a professional
context. We depend upon vendors to provide us with supporting documentation for their claims,
and we rely on our colleagues to help us select the best equipment and devices. Because we do
not have an independent testing laboratory such as the Consumer's Union to help us out, we must
rely—at least to some extent—on endorsements from respected colleagues. For this reason we
might even say that from an ethical point of view we are obligated to provide and use product
endorsements.
Obligated is a strong word and deserves further comment. For that, I turn to the literature of my
faith community. I am sure others have similar sources.
Mishnah Samuel tells a story about two sages who lived 2100 years ago. They were walking in
the countryside when they were confronted by a local resident. The resident questioned the right
of the healer to heal if, as they all agreed, it was God who brought the illness. The sages
responded that just as we are commanded by God to tend trees and crops, so must we tend the
human body. Further, because saving a life is one of the strongest commandments in the Hebrew
Bible, and because curing the ill is linked to this commandment, healing is considered a major
obligation.
It is a short step from the obligation to heal to the obligation to share our knowledge of helpful
healing tools and devices. As physicists we are not empowered to treat patients directly.
However, we do possess knowledge needed by those who are so empowered. We can provide our
best advice only if we are using the best equipment available. We can only know what that
equipment is if we are well informed. Thus, it is essential that those who know be encouraged to
share their knowledge with those who need to know. Further, the sages taught that a physician
who heals for nothing is worth nothing. The meaning, of course, is that fair remuneration is
expected.
505
Physicists have a moral responsibility to lend their reputations to commercial products they
believe in. We need to be sure, however, that we do not abuse that responsibility. That is a truly
fine line to walk, and we should take great care in doing so.
Rebuttal
The truth is, having argued against the proposition, I am actually somewhat hesitant to oppose it
wholeheartedly. The reason is that it is quite difficult, at worst, to remove crass selfinterest from
the picture, or, at best, to eliminate the appearance of a conflict of interest. Yet we still need the
help that endorsements potentially can provide when making purchasing decisions.
This paradox may be characterized by the often-conflicting need for individual freedom, as
opposed to the constant demand for group responsibility to the group. Pursuing the latter need
not exclude the former. However, a paradigm shift in individual members of the group, and in the
group as a whole, may be required for this effort to succeed. Moses Pava1 refers to this shift as
the "logic of appropriateness" rather than the "logic of consequence." In practice, such a shift
involves asking oneself, "how does this decision help the group?" instead of, "what's in it for
me?" The two are not necessarily mutually exclusive.
Organizational guidance in this area is helpful. Studies have shown that most people want to do
the right thing when they are able.2 Pava argues that in order to promote ethical behavior in its
employees, an organization must model that behavior in word and deed. This includes clear
statements of what the expectations of the organization are, and adherence to published ethical
guidelines by the organization's leadership.
Individual members must also participate in organizational efforts to enable ethical behavior. As
Bowen McCoy put it, "What is the nature of our responsibility if we consider ourselves to be
ethical persons? Perhaps it is to change the values of the group so that it can, with all its
resources, take the other road."3 The AAPM Ethics Committee is currently wrestling with this
issue. Participation in discussions at meetings and written input are always welcome.
REFERENCES
1. M. L. Pava, Business Ethics: A Jewish Perspective (Ktav Publishing House, Inc., New
York, 1997).
2. A. Etzioni, The Moral Dimension: Toward a New Economics (Free, New York, 1988).
3. B. McCoy, "The parable of the Sadhu," Harvard Business Rev. 61 (1983).
506
10.23. Medical physicists should not encourage or participate in off-

label, non-FDA-approved uses of intravascular brachytherapy
devices
Shirish K. Jani and Howard Ira Amols
OVERVIEW
FDA-approved medical devices such as intravascular brachytherapy sources may be used for an
off-label (nonapproved) application if, in the judgment of the physician, such use is warranted for
an individual patient. Often an off label use is being evaluated in a sponsored clinical trial, with
the aim of accumulating enough data to determine whether FDA approval should be sought for
the off label use of the device. Physicians frequently are tempted to use a device off-label,
because they believe the device may benefit a particular patient even though the accumulated
data are insufficient to verify the belief. Participation in such use presents an ethical dilemma for
the physicist. This dilemma is explored in this Point/Counterpoint.
Arguing for the Proposition is Shirish K. Jani, Ph.D. Dr. Jani received his Ph.D. in Molecular
Physics from North Texas State University in 1980. After completing a post-doctoral fellowship
at the Medical College of Virginia, he became Chief of Clinical Physics at the University of
Iowa. Since 1993, he has been at the Scripps Clinic in La Jolla, CA. He is certified by the ABR
and the ABMP. Dr. Jani has served on many AAPM task groups and is an active member of the
ACR Commission on Physics. He has served as an ABR Oral Examiner. Dr. Jani is a Fellow of
the American College of Radiology and the AAPM.
Arguing against the Proposition is Howard Ira Amols, Ph.D. Dr. Amols received his Ph.D. in
Nuclear Physics from Brown University in 1974, followed by post doctoral training at Los
Alamos National Laboratory. He has held medical physics positions at the University of New
Mexico, Brown, and Columbia Universities. He is currently Chief of the Clinical Physics Service
at Memorial Sloan Kettering Cancer Center in New York City. He is certified by the ABMP, and
is a Fellow of the AAPM. During his tenure at Columbia University he was intimately involved
in the early development of IVB. For better or worse, he is probably best known for penning
numerous infuriating columns in the AAPM Newsletter.
FOR THE PROPOSITION: Shirish K. Jani, Ph.D.
Opening Statement
An intravascular brachytherapy device approved by the food and drug administration (FDA) for
marketing has clear and well-defined labeling for its use. The device label includes indications
for use, precautions, and radiation dosage. The FDA grants its approval on the basis of safety and
efficacy as established by well-designed clinical trials. For example, a 30 mm long Sr-90 source
train is approved by the FDA for the treatment of restenosis in a native artery where the lesion is
2 cm. If such a device is used to treat a saphenous vein graft (SVG) lesion or a 6 cm arterial
lesion, the application is considered an off-label use. Physicists should neither encourage nor
507
participate in off-label uses of an intravascular brachytherapy device. This opinion is based on

the following three reasons.
First, in an off-label use, the device has not been proven to benefit the patient. The FDA requires
that studies be conducted under carefully-controlled conditions to document such a benefit. It is
unethical to use a device off-label when such use is under investigation at other centers. As an
example, the role of intravascular brachytherapy to treat SVG lesions and long arterial lesions is
currently being investigated in multi-institutional trials.
Second, patient safety is not established for an off-label use. A good example of such concern is
the treatment of a long lesion by multiple stepping of a short FDA-approved source. A recent
publication highlights a case of a late acute thrombosis 15 months after such a stepping
treatment.1 The accompanying editorial points to double-dosing or even triple-dosing the arterial
segments where a deliberate source overlap of 5 mm was used with a Sr-90 device.2 The high
dose may have prevented re-endothelialization, leading to a thrombotic event. The patient pays
the price of such misuse, either in late thrombosis or in terms of years of antiplatelet therapy. By
one estimate, one in four late thrombotic events are fatal.2
Finally, why should physicists be involved with the issues mentioned above? Although we do not
prescribe radiotherapy treatments and doses, we are (and should be) full members of the team
caring for the patient. Physicists are not merely technical staff; they are an integral part of the
professional group involved in the decision-making process. Physicists have an obligation to
institutions and patients to assure safe and effective radiotherapy. This obligation includes not
participating in off-label applications of intravascular brachytherapy devices.
Rebuttal
Dr. Amols correctly points out that off-label use of an IVBT device is not illegal. However, in
the absence of sufficient data, an off-label use is still an experiment on the patient. It should be
done under a protocol and within the framework of the Institutional Review Board (IRB). The
main issue is the patient's right to know whether he/she is being subjected to an unproven
procedure.
I disagree with Dr. Amols' assertion that patients may be in imminent danger if IVBT is not
performed. Even when other options are limited, an investigational protocol can easily be
established. This is precisely what is done at our institution, and certain patients are treated under
a "compassionate use" protocol.
A central question is how physicists can discourage off-label use. First, there needs to be
discussion up-front with the radiation oncologist and the cardiologist on possible off-label use.
As pointed out by Dr. Amols, the physicist needs to evaluate the dosimetric implications of, for
example, sequential stepping of a line source. Second, the physicist may get the IRB involved.
Finally, when physicists are pressured to become unwilling participants, the issue of liability may
be raised. Use of an unproven device can compromise the defense of a medical negligence case.
Overlapping of external fields, resulting in twice the dose to a region of the spinal cord, is
considered negligence. Similarly, over/under dosing of an arterial segment without full
knowledge of its implication is negligence. Physicists cannot hide behind some technicality that
off-label use is not illegal according to the FDA; or that the NRC does not object to such
procedures.
508
I made my argument on scientific/medical merits. Dr. Amols rightfully argued from an

administrative perspective. It is a fact that physicists are professionals who take an active role in
medical care and engage in intellectual discussions about right and wrong choices. But the choice
of off-label use is not simply a choice of right or wrong. Often it is a choice of financial gain over
the medical needs of the patient.
The subject of off-label use is relatively new to the radiotherapy community. As the field moves
to more interdisciplinary applications of radiation, the issue of off-label use will undoubtedly
continue to arise. The stand we take today will set an example for the issues of tomorrow.
AGAINST THE PROPOSITION: Howard Amols, Ph.D.
Opening Statement
First, let us point out that off-label use of an FDA approved device is not illegal. FDA regulates
the manufacture and sale of medical devices, but does not regulate the practice of clinical
medicine, at least not to the extent of dictating its practice by individual physicians. Off-label use
of a medical device is neither illegal nor uncommon. It is perfectly legal for a licensed physician
to treat any part of the human body with radiation. At the risk of hyperbole, we point out that
radiation therapy of acne, athletes foot, baldness, and halitosis are all on-label uses of radiation.
Only intravascular brachytherapy (IVB) of an artery is off label. Treating the fore-mentioned
conditions with radiation may be deemed unwise, or even stupid, but it is not off label, and
certainly not illegal. Neither stupid nor off label are necessarily illegal. One may ask why IVB
has been singled out by the FDA as a "significant risk device," and has precisely defined criteria
for "on-" and "off-" label use, while radiation treatment of conditions such as acne and halitosis is
acceptable. This, however, is a topic for another Point/Counterpoint column.
A principal responsibility of the medical physicist in IVB, as in any radiation procedure, is to

advise the physician on physics, dosimetry, safety, and quality assurance aspects. If a physician,
using the best clinical judgment, decides that the medical needs of a particular patient will be best
served by an off-label use of an IVB device, then the physicist has an obligation to remind the
physician that what is being proposed is off label, and to advise the physician about the possible
dosimetric and safety consequences and uncertainties of the proposed procedure. If, for example,
an off-label use might cause a 100% overdose to a segment of artery, the physicist must inform
the physician of this possibility. But if, knowing these facts, the physician still believes that the
off-label application is in the best interests of the patient (a patient, by the way, who might die of
a heart attack if nothing is done), then the physicist has a professional obligation to assist the
physician in performing the procedure.
If a physician consistently engages in clinical activities that a physicist believes are unwise or
reckless, and consistently ignores the recommendations of the physicist, then a sensible physicist
should start looking for another job—even if nothing illegal has transpired. Only when a
physician's consistently bad judgment and off label use of IVB devices crosses the line from ill-
advised to illegal does the physicist have an obligation to refuse to participate. The physicist then
also has an obligation to report such activities to the proper authorities.
But let's not pretend that we have the medical training to determine when off-label use of a
perfectly legal medical device is or isn't in the best interests of a particular patient. Especially not
509
when we are working side by side with radiation oncologists and interventional cardiologists who
do have that expertise.
Rebuttal
Dr. Jani correctly states that IVB devices are approved for marketing with well defined labeling
for their use. He suggests that physicists should not participate in off-label use. We must,
however, differentiate between approved for marketing, and approved for use. FDA regulates the
marketing of medical devices by manufacturers, but not the use of such devices by licensed
physicians. I also see a distinction between physicists not encouraging off-label use, as opposed
to not participating in off-label use. Sometimes, an experienced physician will conclude that off-
label use of an IVB device is in the best interests of a particular patient suffering from life
threatening cardiovascular disease, especially if the patient has few alternative medical options
remaining.
Many IVB patients have failed (i.e., recurred after) multiple previous interventional procedures,
and the physician may have little choice but to offer them either off-label IVB treatments, or slow
death! Imagining myself as a patient, I would rather have that choice made by my personal
physician rather than by a well meaning regulator in Washington who has never seen my EKG or
angiogram.
Like Dr. Jani, I would never encourage a physician to routinely engage in off-label medical
procedures. But refusing to participate may result in even more harm to some patients. Dr. Jani
suggests that some patients may already have suffered late thrombosis as a result of being
overdosed during off-label IVB procedures. This may be true, but it can only be determined from
controlled clinical trials, which we both support. But even if true, there is still a balance between
risks and benefits for some patients. Physicists have the expertise to assess dosimetry and safety
issues associated with off-label use, and the obligation to counsel physicians on such matters, but
it is the physician who is trained to make the final decision.
Physicians guilty of consistently bad judgment should be barred from practicing medicine. But
let's not limit the treatment options good physicians can offer their patients by having physicists
walking out of the treatment facility. It is not our duty to veto the informed decisions of
physicians to treat particular patients.
Life is full of hard choices, particularly for patients with life threatening cardiovascular disease.
Sometimes off-label use of a medical device is in the patient's best interests. Let's leave such
decisions in the hands of competent physicians who have the benefit of our expert physics
council.
REFERENCES
1. F. Liistro et al., "Late Acute Thrombosis After Coronary Brachytherapy: When is the risk
over?," Cathet. Cardiovasc. Intervent. 54, 216–218 (2001).
2. T. J. linnemeir, "Radiation and late thrombosis: issues, facts, and solutions," Cathet.
Cardiovasc. Intervent. 54, 219–220 (2001).
510
10.24. Medical Physicists should actively discourage institutions

from advertising technologies such as IMRT and HDR
Brachytherapy in order to recruit patients
Marc Edwards and William Bice
OVERVIEW
The acquisition of sophisticated technologies is frequently used as a marketing ploy to attract

patients to healthcare institutions. This strategy can be interpreted as simply an example of
capitalism and free enterprise. However, one can also argue that the approach is exploitive
because most patients have little idea of what the technologies are and whether they might benefit
from them. This controversy is explored in this month's Point/Counterpoint.
Arguing for the Proposition is Marc Edwards, Ph.D. Dr. Edwards is the chief physicist for
Radiation Oncology Associates of Kansas City. During the "academic phase" of his career, Dr.
Edwards participated in training medical physicists, residents and therapists at the University of
Missouri's Columbia and Kansas City campuses. He is a fellow of the AAPM, has served on
committees of the AAPM and ACR and is a former council member of the National Council on
Radiation Protection and Measurements. He is certified in therapeutic and diagnostic radiological
physics by the ABR and also occasionally serves as an oral examiner for the American Board of
Radiology.
Arguing against the proposition is William S. Bice, Jr., Ph.D. Dr. Bice received his doctoral
degree in medical physics from the University of Florida in 1985. Soon after, he was assigned to
the Radiation Oncology Service at Brooke Army Medical Center in San Antonio. After leaving
the military in 1989, Dr. Bice began providing consulting physics services, founding his current
firm, International Medical Physics Services, in 2000. He is certified in therapeutic and
diagnostic radiological physics by the ABR. Dr. Bice holds an adjunct faculty position in the
CAMPEP-accredited medical physics program at the University of Texas Health Science Center
at San Antonio.
FOR THE PROPOSITION: Marc Edwards, Ph.D.
Opening Statement
Medical physicists can be justifiably proud of their role in the development and clinical delivery
of intensity modulated radiation therapy (IMRT) and high dose rate (HDR) brachytherapy. These
advances offer the hope, and accumulating evidence suggests the reality, of improved outcomes
in cancer treatment. The technologies are expensive, and it is important to seek adequate
reimbursement for their use. One response to the economic pressure of high cost is to increase
utilization. However, economic pressure should not subvert the professional and ethical
responsibilities of the healthcare community.
511
Early objections to Direct-to-Consumer (DTC) advertising focused on unsubstantiated clinical

claims for IMRT, together with possible, but as yet unknown, long-term complications.1
Additional discussion introduced issues of "protected personal speech," "restricted commercial
speech" and the role of the Food and Drug Administration (FDA) in advertising regulation.2 This
discussion assumes that the efficacy of IMRT and HDR brachytherapy is no longer in doubt, and
that advertisements are not in violation of the law.
DTC advertising is moderately successful, at least for prescription drugs. A Government

Accounting Office (GAO) study found that in 2001, drug companies spent about $2.7 billion on
DTC advertising, and that 5% of patients requested and received a prescription for a particular
drug in response to DTC advertising.3 The greatest amount of prescription drug advertisement
money is spent on proprietary (i.e., nongeneric) drugs for chronic conditions, which yield a high
return on investment for drug companies. If the reimbursement for IMRT was not many times
more than conventional treatment, one probably would see less advertising.4
Much DTC advertising is justified as promoting patient awareness and public education about a
particular disease and its possible treatments. Thus the patient is "assisted" in making a choice by
being made aware of available options. However, decision-making can be overwhelming in the
context of illness.5 Conflicting and sometimes confusing information is available from healthcare
providers (some in the form of advertising), support groups, books, journals and the Internet. This
overload of information can prevent the patient from making good decisions. The proper role of
the healthcare provider is to facilitate the transformation of information into knowledge for each
individual patient.5 If DTC advertising is represented as altruistic education and awareness, then
the advertising funds should be donated to independent groups, such as the American Cancer
Society, whose mission is in part to provide unbiased information.
The traditional role of the healthcare community, arising from ancient times, was fiduciary, with
protection of the patient's interest the primary concern. Respect for others, empathy, compassion,
honesty, integrity and professional excellence are the roots of medical professionalism. Among
the key values of medical professionalism are service, advocacy and altruism. Business values
such as cost, profit and competition contrast sharply with medical values. The fundamental
problem with DTC advertising is that it corrupts the relationship between the healthcare
professional and the patient by making it captive to both medical and business values. The patient
becomes a "consumer," to be pursued in a marketplace by providers competing to capture
"market share." Health care becomes a commodity to be bought and sold at the market price, and
the patient-consumer must disentangle the economic self-interest of the supplier from the value
of the product. Healthcare providers and patients have protested loudly when managed-care
companies have attempted to impose market dynamics on the patient-provider relationship. Yet,
healthcare providers are, by their actions in directly pursuing patients, self-imposing similar
market dynamics.
The ethical implications of the commodification of health care have been explored by E.D.
Pellegrino, from the Center for Clinical Bioethics.6 In a commodity transaction, the ethics of
business replaces professional ethics. Business ethics legitimate self-interest, competitive edge
and a level of treatment based on the purchaser's ability to pay. When health care is viewed as a
commodity, providers seek to "sell it" like any other commodity, including using advertising to
create demand among those who can pay. In rejecting this model, Pellegrino argues that health
care is not a commodity and that treating it as such is deleterious to the ethics of patient care. He
avers that health is a "human good," and that an ethical society has an obligation to protect it
from a market ethos.6
512
The most important thing to a patient is not the availability of some high technology device,
rather it is the ability of a team of physicians, physicists, dosimetrists and therapists to use a
technology with skill for the benefit of the patient. Such a commitment can be communicated
only within the framework of a professional fiduciary relationship between the healthcare
provider and the patient. Pursuit of patients through DTC advertising detracts from this
relationship. The moral and ethical responsibility of healthcare providers, including medical
physicists, is to discourage marketing activities such as DTC advertising.
Rebuttal
Physicists should not support DTC advertising because it is "here to stay," or because the other
guy does it. Competition, when it is for academic insight or technological advancement, does
ultimately benefit the patient. It is much less evident that advertising to induce a patient to seek
treatment with a particular technology or at a particular center results in better patient care or
societal benefit. DTC advertising does put pressure on the healthcare system, but this pressure is
to respond to a synthesized demand, rather than to advances demonstrated by clinical research
and affirmed by professional standards. Let's affirm our traditional fiduciary responsibility to
patients by rejecting DTC healthcare advertising.
AGAINST THE PROPOSITION: William Bice, Ph.D.
Opening Statement
Advertising in medicine is legal. The Sherman Anti-Trust Act of 1890, with its primary focus on
discouraging monopolistic practices, laid the foundation for advertising by professionals. Since
then United States courts have repeatedly upheld the right of physicians to advertise, largely on
the basis of free speech. In 1975 the American Medical Association was successfully sued by the
Federal Trade Commission for restricting advertising through its code of ethics. It became
unlawful for physicians to restrict medical advertising in 1977. 7
Legal, of course, is not the same as ethical.
The concept of what is ethical in medical advertising has changed dramatically in the last
100 years. Called reprehensible at the turn of the last century, largely ignored except for the
direct solicitation of patients by the 1960s, accepted and overwhelmingly practiced in today's
medical marketplace, the ethical nature of medical advertising has evolved. The AMA now states
that "there are no restrictions on advertising by physicians except those that can be specifically
justified to protect the public from deceptive practices." The entire thrust of the guideline is that
physicians and patients must not be misled. 8
As an example, consider the promulgation in 1997 of guidelines by the Food and Drug
Administration for electronic advertising of prescription drugs. 9 These guidelines were published
to clarify what keeps an advertisement from being misleading. With these guidelines, drug
companies were provided firm legal ground upon which to base their direct-to-consumer
advertising efforts. Many feared the consequences of opening the advertising floodgates,
primarily citing increased prescription drug costs. Rosenthal , found that, while spending for
advertising prescription drugs has increased dramatically from 1996 to 2000, as a percentage of
overall sales the advertising cost has remained relatively constant (at about 14%). Additionally,
the cost for direct-to-consumer advertising, while increasing, has remained at about 1–2% of the
513
sales budget while promotion to healthcare professionals has accounted for the other 12–13%.10
What this means is that advertising probably sells prescription medicines but does not, by itself,
increase the cost.
It is important to remember that the physician is still the gatekeeper. Whether for diagnosis or
therapy, patients are never treated without a prescription. Returning again to the example of
prescription drugs, Rosenthal noted that, although 25% of patients seen by a physician mentioned
a drug that they saw advertised, only 6% actually received a prescription for the drug.10 Doctors
do pay a price for this gate keeping. They cite increased time spent in consultation as the primary
drawback to an increasingly informed patient. Dealing with misinformation and its correlate,
unrealistic expectations, burdens the doctor-patient relationship. Nevertheless, responsibility for
inappropriate prescribing with subsequent increased costs rests upon the physician's shoulders.
An argument can be made for medical advertising based simply upon its prevalence. Television,
radio and newspaper ads promoting healthcare products and services have become ubiquitous.
These ads are no longer restricted to elective services such as plastic surgery or cosmetic
dentistry. Everyone does it. From the sole practitioner promoting his practice on a thinly-
disguised National Public Radio "announcement," to recurring advertisements for prostate cancer
treatment appearing daily in the sports section of the local newspaper, the promotion of medical
services to patients has become firmly entrenched. However seedy these marketing practices may
seem, they are here to stay. And, in a competitive medical marketplace filled with cronyism,
managed care, and physician ownership of treatment facilities, healthcare promotion often is a
matter of survival.
This competition has been good for medicine. Heightened public awareness, more involvement
by the patient (how and where can I get the best treatment?), better patient compliance and a
pressure on providers to offer improved care are all the result of advertising.
Rebuttal
Dr. Edwards, in his eloquent and impassioned opening statement, presents arguments based upon
premises that do not bear up under scrutiny. His assertions fall into three areas: a distinction
between medical and business ethics, the biased nature of direct-to-consumer (DTC) advertising,
and the innate evil in commodification of medical services.
In distinguishing between medical and business values, Dr. Edwards listed some noble notions
indeed! Falling on the side of the medical profession are "respect for others, empathy,
compassion, honesty, integrity and professional excellence," as well as "service, advocacy, and
altruism." These are in contrast with those dastardly business values, "cost, profit, and
competition." It does not follow that making a profit precludes having any or all of the attributes
above. In fact, the case is more likely the opposite. Successfully competitive businesses—the
ones that demonstrate long-term profitability—are those that engender those very same values
that Dr. Edwards ascribes to the medical profession. With regard to the appropriateness of using
advertising as a tool to promote awareness, the crime is not in the promotion of the product or the
procedure, but in using advertising to mislead.
There is an inherent bias in DTC advertising. Dr. Edwards would contend that this bias makes
advertising promotional rather than educational. While the goal of promotion may be different
than that of education, the two are inseparably entwined. There is no such thing as pure
education; it is always presented with a bias. Consider the too frequent example of the patient
514
that has been "educated" by a urologist with regard to the treatment options for early stage
prostate cancer but has never heard the words "brachytherapy" or "radioactive seed implant" or
"IMRT." Education, or at least awareness, accompanies promotion. And, as illustrated in the case
above, the awareness driven by promotion can provide a check on biases from other, more
traditional, sources of patient information. And what if the patient is uneducable? What if the
stresses accompanying serious illness result in "bad" decisions? The physician is still the
gatekeeper; without a prescription, the treatment does not occur.
The practice of medicine has clearly become a business. Medical services are a commodity to be
bought and sold. The rise of HMOs, physician ownership of treatment facilities and negotiated
pricing with insurance carriers attest to this. Today's medical practitioner is faced with market
decisions as confounding as medical decisions. Even in a system with set base-pricing—
Medicare rates—there are competitive pressures to acquire patients. More patients mean more
income, or better equipment, or newer facilities.
Forgotten in all of this commodification is the true consumer, the patient. And like buyers
everywhere, they must beware. Gone are the days, thankfully, when the only input that the
patient had in purchasing medical care was "Well, whatever you think, doc." The patient should
have as much access to as much information from as many sources as he desires. Only then can
he make the grave decisions that are his to make. Across the street from where I live is a
billboard advertising one of the local healthcare systems. A smiling face beams down from this
lofty perch, just above a caption proclaiming "My doctor gave me a choice." Our patients deserve
nothing less.
REFERENCES
1. E. Glatstein, "The return of snake oil salesmen," Int. J. Radiat. Oncol., Biol., Phys. 55,
561–562 (2003).
2. E. C. Halperin, "Free speech and snake oil," Int. J. Radiat. Oncol., Biol., Phys. 56, 1507–
1509 (2003).
3. Government Accounting Office, "Prescription drugs: FDA oversight of direct-to-
consumer a advertising has limitations," GAO-03-177 (2002).
4. E. Glatstein, "Letter to editor re snake oil," Int. J. Radiat. Oncol., Biol., Phys. 56, 1507–
1509 (2003). K. Berger, "Informed consent: information or knowledge?," Med. Law 22, 743–750
(2003).
5. E. D. Pellegrino, "The commodification of medical and health care: the moral
consequences of a paradigm shift from a professional to a market ethic," J. Med. Philos. 24, 243–
266 (1999).
6. The Ethics Committee of the American Academy of Otolaryngology—Head and Neck
Surgery, Otolaryngol.-Head Neck Surg. 115, 228–234 (1996).
7. American Medical Association Ethical Guideline E-5.02, "Advertising and Publicity,"
http://www.ama-assn.org/ama/pub/category/8348.html (July 17, 2000).
8. 21 CFR §202.1 (Revised Apr. 1, 2003).
9. M. B. Rosenthal, E. R. Berndt, J. M. Donohue, R. G. Frank, and A. M. Epstein,
"Promotion of prescription drugs to consumers," N. Engl. J. Med. 346, 498–505 (2002).
515
10.25. Medical physicists in developed countries should actively help

medical physicists in developing countries
Gary D. Fullerton and Joseph Ting
OVERVIEW
Every medical physicist has benefited from the guidance and counseling of more senior members
of the profession. As a result, the profession of medical physics has evolved to a level where it
makes substantial contributions not only to patient care, but also to the ongoing development of
new tools and techniques to improve medical diagnosis and treatment. Each of us in the field has
an obligation to partially repay the debt we owe our mentors by helping others who need
guidance and counseling. Nowhere is the need for help greater than in developing countries in
which medical physics and high-tech medicine are just getting started. However, each of us has
responsibilities to the institution that pays our salary and expects its patients and clinicians to
benefit from our knowledge and expertise. Providing assistance outside the institution takes time
and energy away from those inside the institution who need our attention. Two different
viewpoints on these conflicting demands on our time and effort are presented in this issue of
Point/Counterpoint.
Arguing for the Proposition is Gary Fullerton, Ph.D. Dr. Fullerton is the Malcolm Jones
distinguished Professor of Radiology and Director of the Graduate Program in Radiological
Sciences at the University of Texas Health Science Center at San Antonio. Dr. Fullerton has
served the AAPM on many committees, as Secretary from 1982–1984 and President in 1991.
Currently he is Editor of the Journal of Magnetic Resonance Imaging, Secretary General of the
International Organization for Medical Physics and Secretary General of the International Union
for Physical and Engineering Sciences in Medicine. He is certified in radiological physics by the
ABR and is a strong supporter of the development of the medical physics profession in the
international arena.
Arguing against the Proposition is Joseph Ting, Ph.D. Dr. Ting is Associate Professor in the
Department of Radiation Oncology at the Emory University School of Medicine. He is certified
by, and a frequent oral examiner for the American Board of Radiology. He has authored or co-
authored several articles on the technical aspects of external beam therapy, and has three patents
pending. He teaches both residents in radiation oncology and graduate students in medical
physics, and has been a guest speaker at several institutions and conferences, including Sun-Yat-
Sen Hospital in Taipei and the International Conference on Medical Imaging, Medical Physics
and Precision Radiation Therapy in Guangzhou.
FOR THE PROPOSITION: Gary Fullerton, Ph.D.
Opening Statement
516
The fundamental but oblique question embedded in this issue is, ‗‗Are medical physicists
professionals charged with the development and maintenance of knowledge related to their
practice or are they laborers with duties related solely to the hourly work and wage they
receive?‘‘ Professionals from all walks of life weigh the ethical questions concerning the
allocation and application of time to achieve the greatest good for clients. Laborers, on the other
hand, need only apply their efforts to the allotted task in the contracted time. Medical physics is,
in my view, a profession for which the participants have responsibilities far beyond those of the
given workday task.
Presently there are 4500 medical physicists in the USA and more than 16 500 worldwide that are
members of the International Organization for Medical Physics. For every medical physicist in a
developed country there are two or three practicing in developing countries under less
technically advantaged circumstances. There are many reasons that physicists from developed
countries should focus on assisting these colleagues. These reasons range from enlightened self-
interest to a global concern for humanity.
The ability of medical physicists to practice and provide the diagnosis and treatment of patients
depends on two factors: (1) a specialized knowledge of the field that allows safe and efficacious
patient care, and (2) a set of well designed and maintained technical tools that make their work
possible. These tools range from accelerators to imaging devices and measurement instruments.
If the education and training of medical physicists in developing countries are not over time
brought to the level of those in the developed countries, then medical physics stands to lose.
First, the larger numbers of medical physicists practicing in reduced circumstances are in the
majority. The reduced circumstance could become recognized as the worldwide standard for the
level of medical physics practice to the loss of medical physicists everywhere. Second, the
inadequate application of advanced technologies in developing countries and resultant failures
could undermine public confidence in medical physics techniques. Such losses could make high
technology medicine politically unpopular and treatments less accessible to patients. This places
the future viability of our profession in jeopardy. Medical physicists should defend the integrity
of their profession, if they wish to be respected by the patients they serve as well as by other
medical professionals.
Evolution and progress are key elements in the practice of medical physics and other components
of high technology medicine. The standard of practice today will not be the standard of practice
five years from now. Medical physicists must continue to provide new and better ways to
diagnose and treat disease or be prepared to turn over their profession to less demanding roles
filled by technology specialists. The cost of research and development of new devices remains a
major component in the cost of medical physics and the medical specialties that use the physics
devices. Growing demands for cost effectiveness require that medical physicists spread the cost
of future improvements over a greater fraction of the world population. Medical physics needs to
seek the advantages of globalization just as do other purveyors of high technology solutions to
human problems.
The application of physics to health care is our profession. The medical physicist should make
the fruits of his or her labor available to humanity. Extending commitments to the developing
world preserves the future of medical physics from the rush to improve day-to-day clinical
productivity. The difficulty of the decision typifies the ethical dilemma of all professionals.
Rebuttal
517
Dr. Ting and I agree, ‗‗It is inappropriate for physicists to toot around the globe.‘‘ We differ in
my belief that there are many legitimate reasons for American medical physicists to participate in
international conferences, cooperative research projects and cooperative educational programs.
The professional medical physicist needs to assess benefits and costs to reach an ethical decision
concerning international participation. The example of the meetings in China discussed by Dr.
Ting is a good one. When I visited China in 1990 there were only 20 MR units in the country;
now there are thousands. Chinese patients profit from both the technology and the clinical
interpretation skills gathered from the international literature. The idea that IMRT and EPID can
be implemented in China, together with the interpersonal interchanges (friendships) that are
being developed with the inventors of these techniques, are motivators for change in China. In
1990 China did not have a society of medical physics; today China is a member of the IOMP
with more than 400 individual members. In addition, Taiwan and Hong Kong medical physics
societies report 150 and 50 members, respectively. My graduate program in San Antonio now
receives applications for medical physics graduate study from China that include masters degree
training in medical physics at Chinese universities. The meetings that Dr. Ting dismisses for low
productivity have been potent harbingers of change in China.
AGAINST THE PROPOSITION: Joseph Ting, Ph.D.
Opening Statement
In recent years, health dollars allocated to patient care in the U.S. have been drastically curtailed.
Often, costs of radiation treatments or diagnostic procedures far exceed the dollars recovered
from third party payments. Most hospitals and clinical departments are operating under stressed
budgetary constraints with no end in sight. With these constraints in mind, it is inappropriate for
medical physicists to toot around the globe and support medical physics activities in foreign
lands.
A medical physicist who attends a foreign meeting in support of medical physics activities costs
his or her employer approximately $1,000 per day in salary and benefits. Lodging,
transportation, and other incidental costs should also be added. Medical physicists who wish to
provide aid to foreign countries should obtain separate and dedicated funding from other,
government or private, sources.
The total direct and indirect labor costs plus expenses exceeded one million dollars for the U.S.
physicists who attended the two most recent medical physics meetings in China (October 1999
and May 2000). Most of these physicists were not reimbursed by the conference hosts for their
expenses or time and effort. Instead, U.S. institutions generously donated one million dollars to
those conferences in the form of labor and direct expenses.
I am not suggesting that U.S. medical physicists should not contribute to training foreign medical
physicists who will function in important roles in their countries. I have personally trained many
foreign medical physicists who have later become important contributors either in America or in
their home countries.
I do object, however, to supporting and hosting conferences in foreign lands. Meetings and
lectures offer high visibility but low productivity and they are the wrong forum for training and
teaching purposes. For example, how many Chinese hospitals and patients will benefit from the
IMRT and EPID presentations and discussions that occurred during the two China meetings
mentioned earlier?
518
There are many aphorisms cautioning us to spend money wisely; for example: ‗‗A penny saved
is a penny earned‘‘ and ‗‗there is no free lunch.‘‘ Most of us draw our salaries from our hospital
or clinic employers that, in turn, are supported by third party carriers, and indirectly, patients.
There is no provision in this support to underwrite medical physics activities in foreign lands.
Medical physicists in the U.S. should spend their time attending to challenges on our home front
and not compromising their efforts by focusing on medical physics activities in foreign countries.
There are more cost-effective ways than meetings to aid foreign medical physicists and
strengthen medical physics programs abroad. They include, for example: Sponsoring foreign
medical physicists to spend time at U.S. hospitals; establishing ongoing working relationships
with foreign hospitals; formulating joint training sessions; teleconferencing; and providing grants
for foreign medical physicists to attend special purpose workshops (not general meetings).
Rebuttal
I have known Dr. Fullerton for over twenty years and I have a great respect for his professional
and scientific endeavors. However, assumptions and statements made in his ‗‗Opening
Statement‘‘ are false. Here are a few examples:
(1) ‗‗Professionals‘‘ are not the only persons who need to consider the allocation of time and
effort to achieve the greatest good for clients. ‗‗Laborers‘‘ who receive hourly wages need to do
the same. I do not understand Dr. Fullerton‘s differentiation of ‗‗professionals‘‘ and ‗‗laborers.‘‘
We all are workers to achieve the greatest good for our clients. In this case, our ultimate clients
are patients.
(2) I do not believe that there are 16 500 medical physicists who are practicing medical physics
in a manner directly impacting the well-being of patients and their diagnostic and treatment
outcomes. I do not argue against physicists sharing knowledge with others. But, spending money
hosting conferences abroad is not an efficient method to provide aid. Dollars could be better
spent by sponsoring foreign medical physicists to study at institutions of excellence in this
country.
(3) The danger of lowering the standard of practice is far fetched. I often tell my children to
‗‗look up and never down.‘‘ Standards of practices will continuously improve and cannot be
‗‗reduced‘‘ as Dr. Fullerton imagines. It is human nature. Otherwise, we would still be living in
caves and cooking with twigs.
(4) I do not challenge the desirability of a mutual exchange of ideas and inventions with persons
from abroad. But sponsoring and attending conferences are not effective forums for such
exchange.
Finally, we should stop wasting precious dollars sponsoring and attending meetings abroad. We
should not use the excuse of ‗‗educating medical physicists‘‘ to get a paid trip abroad. We should
take ‗‗educating foreign medical physicists‘‘ more personally and carefully. It is a very serious
commitment.
519
10.26. The World Wide Web is a net disservice to medical physicists

in developing countries
Kwan-Hoong Ng and Azam Niroomand-Rad
OVERVIEW
Many people, physicists among them, believe that the World Wide Web (WWW) is a great
service to developing countries because it provides access to information about sophisticated
healthcare technologies and methods. Some believe, however, that the WWW is a disservice
because it raises hopes and expectations, and encourages deployment of sophisticated equipment
and procedures that require resources that would be better used to support fundamental public
health and education efforts. These opposing viewpoints are the topic of this issue of
Point/Counterpoint.
Arguing for the Proposition is Kwan-Hoong Ng, Ph.D. Dr. Ng is Professor and director of the
medical physics program at the Department of Radiology, University of Malaya Medical Centre,
Kuala Lumpur, Malaysia. He obtained his master's degree in medical physics and biomedical
engineering from the University of Aberdeen United Kingdom and Ph.D. in medical physics
from the University of Malaya. He is one of those rare physicists outside the US who is board
certified by the American Board of Medical Physics. Dr. Ng is the editor and co-founder of the
Electronic Medical Physics World (EMPW). He serves on several committees in the
International Organization for Medical Physics (IOMP) and a council member for the
International Union of Physical and Engineering Sciences in Medicine (IUPESM). He is the
founding president of the South East Asian Federation of Medical Physics. Dr. Ng has published
over 60 papers in peer-reviewed journals.
Arguing against the Proposition is Azam Niroomand-Rad, Ph.D., an Iranian by birth. Dr.
Niroomand-Rad received her Ph.D. in Physics from Michigan State University in 1978. She
completed her Medical Physics postdoctoral training at the University of Wisconsin —Madison
in 1983. She has received an Honorary Doctor of Science from Cazenovia College, N.Y., in
2001. She is now a Professor and Clinical Physics Director in the Department of Radiation
Medicine at Georgetown University Hospital in Washington D.C. She is board certified by the
ABR in Therapeutic Radiological Physics and by the ABMP in Radiation Oncology Physics. She
is a Fellow of AAPM and has served on many AAPM committees, task groups, Board of
Directors (twice), as well as President of the AAPM Mid-Atlantic Chapter. She has served IOMP
as Chief Editor for Medical Physics World, and has been active on many IOMP/IUPESM
committees. She is now serving AAPM as Associate Editor for Medical Physics, Chair of
International Affairs, and Chair of International Scientific Exchange Programs where she has
organized radiation therapy courses/workshops for many developing countries. She is presently
the Vice-President of the IOMP.
FOR THE PROPOSITION: Kwan-Hoong Ng, Ph.D.

520
Opening Statement
The advent of the WWW makes it easy to locate and retrieve information. Important medical
developments and discoveries can be disseminated expeditiously throughout the world, and most
medical professionals have rushed to embrace this technology, hoping to reap the benefits. But
could the World Wide Web actually be a disservice to medical physicists in developing
countries?
The gulf between the plugged-in and the not plugged-in world is widening. Developing countries
struggle with the high cost of a basic electronic infrastructure, which is a precondition for
enjoying the benefits of any high technology, including the WWW. It was reported recently that
wealthy industrial nations, comprising less than 20% of the world's population, account for 80%
of Internet users.1 This disparity exists because the cost of being connected is prohibitively high
in most developing countries. In the United States, for example, the monthly internet access
charge represents a mere 1% of the average monthly income. In Sri Lanka, Bangladesh and
Nepal it is 60%, 190%, and 280%, respectively. Accessing and downloading data-intense
information requires high bandwidth and a stable communications network. The total bandwidth
for all of Latin America is roughly equal to that of Seoul, Korea. All of these issues greatly
handicap medical physicists in developing countries who would like to take advantage of the
WWW.
The challenges facing medical physicists from developing countries differ greatly from those
encountered by their counterparts in advanced countries, because the access to technology,
socioeconomic landscape and other factors vary. In developing countries, basic sanitation, safe
drinking water, and basic health services are major needs. Scarce public funding, if available at
all, is and should be channeled to solving these fundamental problems, rather than on providing
access to high-tech medical know-how and medical physics services that would benefit only a
small portion of the people.
Medical physicists in developing countries, especially younger physicists, have the

misconception that instant answers to their queries can be obtained from the WWW. They do not
realize that basic knowledge can be mastered best by studying standard textbooks, reading
journals, and performing hands-on experiments. They become disillusioned when the hospitals
they work in are unable to purchase equipment necessary for the performance of tasks gleaned
from the web. This disillusionment contributes to the brain drain of medical physicists from
developing countries, adding to the many setbacks faced by these countries.
Although medical physics list-servers have opened up opportunities for global communication,
the selection and discussion of topics are dominated by physicists in developed countries, and
their relevancy to the needs of developing countries is questionable. Considerable time is needed
to sift through these postings, and one can become confused and disorientated at the end of the
day, particularly because scientific facts are not distinguished from personal opinions. Medical
physicists from developing countries often feel intimidated about posing questions, because their
questions have attracted sarcastic remarks in the past. This adds to their frustration.
In conclusion, we cannot deny that the web is a net disservice to medical physicists in developing
countries.
Rebuttal
521
Admittedly many journals offer free online access and content alert services. However, these
services tend only to tantalize the appetite of the readers because they are unable to access full
articles from the two leading journals, Medical Physics and Physics in Medicine and Biology.
There is no reduced subscription cost to these journals for online access only. My colleague
claims that the global medical physics listserver allows physicists in developing counties to
communicate with others in advanced countries. However, only a mere 4% of subscribers to the
listserver are from developing countries. Most physicists in developing countries cannot afford
access to the WWW. How can the listserver be described as a total service when it does not reach
a large number of physicists?
Free distance learning and other free stuff are not free, because one has to pay for access and,
often, for enrolling. And virtual applications have major limitations that do not measure up to
hands-on experience for basic skill acquisition. Competency is acquired from working with
experienced and knowledgeable medical physicists, not through self-learning from the WWW.
The investment necessary to implement new clinical protocols and establish the efficacy of new
techniques is expensive. Most developing countries still lack basic equipment and test tools.
If the WWW has such a positive impact on teaching and educating medical physicists in
developing countries, then why are organizations such as the American Association of Physicists
in Medicine (AAPM), International Atomic Energy Agency (IAEA), and International
Organization for Medical Physics (IOMP) sponsoring and funding workshops and courses for
developing countries?
My colleague questions how the WWW could have a negative influence in any developing
country. The focus is not on negative influence, but rather on whether the WWW is a total
service or a disservice. It is my opinion that the WWW is a net disservice.
AGAINST THE PROPOSITION: Azam Niroomand-Rad, Ph.D.
Opening Statement
In my opinion, the WWW has already had a very positive impact and will continue to serve
medical physicists worldwide, particularly in developing countries. The development and impact
of this technology in the teaching and education of medical physicists, and in the dissemination
of information to medical physics communities, will be substantially different in the coming
years. The amount of scientific information on the web is large and will continue to grow. One of
the major developmental needs is to organize the information so that it can be located easier and
faster. I believe that the net benefit from this technology will be even greater for medical
physicists in developing countries than for medical physicists in developed countries.
Like all new technologies, developed countries accommodated to WWW technology earlier than
developing countries. But developing countries have benefited more from the WWW, by
combining an old technology—the telephone— with the computer. The WWW permits
worldwide dissemination of scientific knowledge and rapid communication among individuals
and groups. It has facilitated regional and international interactions, and thus participation of
medical physicists from developing countries in virtual committee and council meetings without
enduring (often unaffordable) travel expenses. As an example, medical physicists in developing
countries can communicate with other medical physicists by joining the global medical physics
list server ([email protected]) and they can use the "Ask your Medical Physicist"
522
(www.medphysics.wis.edu) site to direct any medical physics question to experts. These

excellent services, which have been produced by a few medical physicists, are free of charge.
In addition, WWW technology has allowed free dissemination of scientific and professional
information, free distance learning, and free access to abstracts and published papers to medical
physicists worldwide. For example, the IOMP has taken a lead in posting medical physics
information and activities on the www.iomp.org website. One of the IOMP long-term goals is to
post information about all the medical physics graduate programs world wide (including the core
curriculum), minimum required standards for medical physicists, and codes of practice in various
aspects of medical physics. AAPM refresher courses are available on the web (www.aapm.org),
again free of charge. There are some electronic journals such as the Red Journal (International
Journal of Radiation Oncology, Biology, Physics) and Green Journal (Radiotherapy and
Oncology) that are available electronically to all medical physicists free of charge.
It is hard to understand how the WWW could have a negative influence in any developing
country, when all this information is made available to everyone without enduring much
economical hardship. A difficulty in some developing countries, of course, is the lack of English
fluency, which cannot easily be overcome because the availability of scientific materials in most
other languages is so limited.
Rebuttal
Dr. Ng argues that the gap between the plugged-in and the not plugged-in world is widening
because the cost of being connected is prohibitively high in most developing countries. He
defends his position by computing the average monthly internet access charge, as a percentage of
average monthly income for individuals living in the USA, and comparing it to similar
calculations for physicists in Sri Lanka, Bangladesh, and Nepal. In my opinion, this comparison
is not valid because it is not necessary to own a personal computer with personalized internet
access in order to utilize www technology in developing countries. Considering the many
challenges facing developing countries, as described in Dr. Ng's statement, it is more beneficial
to make proper allocation of public funds to meet the basic health needs of the public.
Widespread utilization of WWW technology should be financed mostly by the private sector for
the benefit of the general public in both developed and developing countries. Electronic
infrastructures should be set up in universities, schools, hospitals, and public libraries to narrow
the seemingly widening gap between plugged-in and not-plugged-in nations. I believe the
existence of WWW technology in developing countries has made governments realize that they
need to provide PCs and on-line access to many organizations if they want their countries to be
connected to the modern world.
I agree with Dr. Ng that textbooks, journals, and hands-on experiments are the best methods to
acquire the basic knowledge and skills needed by a medical physicist in any country. However, I
disagree with his conclusion that utilization of WWW technology by medical physicists has led
to a major brain drain from developing countries. There are many contributing factors to the
brain drain for such countries, including lack of social, religious, and political freedom and
economic hardship. The disillusionment of medical physicists enhanced by the WWW is hardly
the source of the "brain-drain" problem.
I also agree with Dr. Ng that it is not easy to sort facts from personal opinions on the Internet.
Hopefully, this problem will be resolved eventually by development of more efficient search
engines for the WWW. Finally, it is understandable that some medical physicists from
523
developing countries feel intimidated about participating in public discussions posted on the
medical physics list server. In my opinion, those who privately contact medical physics experts
for advice and clarification of scientific and professional issues, greatly outnumber medical
physicists from developing countries who are intimidated by the medical physics list server.
REFERENCE
1. Human Development Report 2001, United Nations Development Programme (Oxford

University Press, New York, 2001), www.undp.org/hdr2001
524
10.27. Results of publicly-funded scientific research should be

immediately available without cost to the public
Colin Orton and Christopher Marshall
OVERVIEW
The Public Library of Science offers to publish (for a fee) scientific articles electronically for
immediate access upon acceptance. Recently, the U.S. House of Representatives Appropriation
Committee called for immediate electronic access to accepted scientific manuscripts describing
research funded by the NIH, possibly through a government electronic repository or
PubMedCentral. There are legitimate societal issues associated with providing immediate access
to results of NIH-funded research. But there are pragmatic issues related to immediate access,
including the potential impact on society memberships, advertising revenues, and participation in
scientific meetings. These issues are debated in this month's Point/Counterpoint.
Arguing for the Proposition is Colin Orton, Ph.D. Dr. Orton obtained his Ph.D. in radiation
physics from the University of London in 1965 and subsequently worked as chief physicist at
NYU Medical Center, Rhode Island Hospital, and Detroit Medical Center, with academic
appointments at NYU, Brown University, and Wayne State University. He is currently Professor
Emeritus at Wayne State and President of the International Union for Physical and Engineering
Sciences in Medicine. Dr. Orton is a Past-President of the AAPM and served as Editor of
Medical Physics from 1997–2004.
Arguing against the Proposition is Christopher Marshall, Ph.D. Dr. Marshall received his Ph.D.
from the University of London, England. He joined the NYU School of Medicine faculty in 1969
where he is now a Professor of Radiology. In 1970 he founded and continues to direct the
Department of Radiation Safety at NYU, which provides radiation safety and diagnostic medical
physics services to NYU, its Medical School and three affiliated hospitals. Dr. Marshall has been
closely associated with the business affairs of Medical Physics for two decades, as Chair of the
AAPM Publications Committee, Chair of the Presidential Committee that established the current
business model and currently as the Chair of the Journal Business Management Committee.
FOR THE PROPOSITION: Colin Orton, Ph.D.
Opening Statement
Scientific journals play a vital role in the dissemination of the results of scientific research,
especially those journals with high standards and peer review of manuscripts. I would never
support any action that might compromise the ability of peer-reviewed journals to continue
publishing with such high standards.
The real question to be addressed in this Point/Counterpoint debate is: "can publicly-funded
research be published with "free access" without causing financial stress to journals that disrupts
525
their ability to provide proper peer review?" I believe the answer to this question is "YES," and I
will support my answer by using Medical Physics as an example.
Medical Physics relies on three major sources of income: the dues allocation from AAPM
members, non-member (mainly library) subscriptions, and advertising, both in the hard copy (the
majority) and the online journal. If we were to allow free online access to all articles immediately
after publication, most of these sources of income would dry up: there would be no reason to
subscribe and no market for hard copies of the journal where most of the advertising revenue
derives. Clearly this is untenable. But what if only a fraction (about one-third, which is the
current proportion of articles in Medical Physics that are publicly funded) of manuscripts were to
be provided "free"? Most of the papers would still be available only to subscribers and, in my
opinion very few subscriptions would be "lost." Indeed, if readers could be induced to visit our
online journal in order to read these "free" papers, I can envisage an actual increase in online ads
due to the increased traffic.
The NIH will now put the author's final version of each manuscript in PubMedCentral subject to
a delay of as long as 12 months after publication. This will not be the same as the published
version of the paper since, for example, it will not include the journal page numbers needed for
referencing the article. However, the NIH is willing to publish a link to the online version of the
manuscript on the publisher's website. The way I see this program working to our best advantage
would be for the AAPM to provide the NIH with "free" links to the full text of all government-
funded research papers in the online journal immediately upon publication. Thus any individuals
who are not subscribers to the hard copy journal but want to read these papers will have to visit
the online journal, where they will not only be subjected to our online ads but will also be offered
access to the Abstracts of all of the other manuscripts, including those for which the full text is
not accessible "free." I suspect that many of these readers will want to view the full text of some
of these papers and will decide to pay to view them. This will provide increased revenue for the
journal. Organized properly, the NIH program could benefit not only the NIH and the public, but
also the AAPM.
Rebuttal
Thanks to the excellent leadership of the AAPM Journal Business Management Committee by
my colleague Dr. Marshall over the past half-dozen-or-so years, our journal is in outstanding
financial shape. Dr. Marshall clearly understands the business of running a scientific journal and,
if the Proposition of this Point/Counterpoint debate related to free access for all publications of
scientific research, then I would agree 100% with all the arguments presented against it in Dr.
Marshall's Opening Statement. Such open access would, indeed, devastate the peer-review
process and force us to institute page charges for authors, which I would agree is not a good
business model. However, this is not the Proposition under debate. Here we are discussing open
access to only publicly-funded research papers which, for Medical Physics, relates to only about
one-third of all articles published. The remaining two-thirds of papers would still be available
only to subscribers. What we have to guard against, of course, is any expansion of open access to
more than just publicly-funded research papers. However, I believe that copyright laws will
protect us from this so we should not oppose the current NIH initiative because of fear that this
might lead to expansion. Instead, we should embrace the initiative as a means to enhance the
readership of our journal.
As I demonstrated in my Opening Statement, because the NIH platform (PubMed) will include a
direct link to each manuscript on the website of the journal,1 the NIH-proposed program should,
526
if handled carefully, lead to an increase in readership, subscriptions, and advertising. This hence
could be a "boon" for the journal and the AAPM, which we should all support. More profit for
the journal means more educational and other programs that the AAPM can provide to members.
We might even see a reduction in our membership dues!
AGAINST THE PROPOSITION: Christopher Marshall, Ph.D.
Opening Statement
The proposition that the results of scientific research might be made available without cost to the
public is a chimera. There is a cost to prepare and disseminate such information, which
ultimately must be born by the public indirectly through taxes or the added cost of goods and
services. Publication on web-based platforms makes content generally accessible through the
Internet, thus creating a demand for "open access"—without a subscription or specific payment.
However, open access potentially undermines the system that gives credibility to scientific
research. While authors may provide free access to their results by posting them on the web at
any time, publication in a peer-reviewed journal provides validation commensurate with the
stature of that journal.
The rigors of peer review, revision and resubmission coupled with publishing standards and copy
editing creates a distinctly new product with added value that is traditionally protected by
copyright. There is significant associated cost and publishers bear other costs for online
publications: to develop and deploy new technologies; to connect content through active links to
references in other publications; to actively manage an archive of published articles essentially in
perpetuity. These costs are quoted to range from $800 to $3750 per manuscript 2—and are larger
for the most prestigious publications because of the extra cost of reviewing and rejecting the
material never published. Libraries and individuals traditionally pay for these services by
purchasing subscriptions. For international journals published in the USA, like Medical Physics,
many subscriptions are from outside the USA, thus spreading the cost.
Since open access eliminates the need for paid subscriptions, how will it be funded? The
proponents complain that commercial journals make profits and the professional society journals
return income to their societies. While this is true, the elimination of profit would eliminate the
incentive to publish and weaken the professional societies and their programs, 3 but the editorial
costs would remain. Most scientific publications are not supported by advertising income, and for
online publications this has limited potential because of their format.
The conclusion is that the authors must therefore pay to publish, 4 which assumes that all are able
to pay or be prepared to subsidize those unable to pay, and that the ability to pay will not bias the
rejection rate. While there are good examples of journals that use the author-pay model,5 these do
not represent a general business model for most publications. 6 Unfortunately, the issue has
become politicized7 and there is pressure for open access regardless of the consequences.8 For
NIH-funded research, resulting peer reviewed manuscripts are to be made freely available on an
NIH platform at the public expense,9 notwithstanding the fact that the peer review process was
conducted at the expense of another publisher. The situation is evolving and the only certainty is
that the public will ultimately bear the cost.
Rebuttal
527
I agree that Medical Physics has multiple sources of income, but for this reason it is not a general
example—many scientific publications depend entirely on subscription income. The premise that
the only threat is from the NIH must also be questioned. Other agencies in the USA and
elsewhere will potentially follow this example and claim the right to redistribute content without
payment or charge. However, the NIH example illustrates other dangers. U.S. law prohibits the
NIH from funding politically incorrect research on stem-cells,10 so the NIH will not provide
automatic open access to associated papers. Is work that has not been funded by public sources
less relevant to the public interest? Seminal work by Darwin and by Einstein does not pass this
test, but is of the highest level of public significance.
The open access movement advocates for open access to all the important research that is
published. The former Director of the NIH, Harold Varmus has urged "young scientists to
consider publishing their best work in open access journals."11 But since some open access sites,
such as those of the Public Library of Science, depend on author fees drawn from research funds
(in addition to grants and donations)12 while others depend on direct public subsidies as in the
NIH model, will politics intrude further into science and publishing through control of funding?
Can we also have confidence that the NIH and others will adequately maintain their open access
sites as politics and the economy influence budget priorities?
Consider the following: All references that I have cited are links to open access websites. They
will only remain accessible if these sites are maintained—long term archival access requires
motivation and economic stability, which has been traditionally built around the subscription
model of publishing. Can we risk literally giving this away?
REFERENCES
1. Policy on Enhancing Public Access to Archived Publications Resulting from NIH-

Funded Research. NIH Notice Number: NOT-OD-05-022 (February 3, 2005);
http://grants.nih.gov/grants/guide/notice-files.
2. Association of Learned and Professional Society Publishers Response to Science and
Technology Committee Inquiry into Scientific Publications, p. 7 (2004);
http://www.alpsp.org/2004pdfs/S&TcommALPSPresponse.pdf
3. F. Gannon, Ethical profits from publishing, EMBO reports 5, 1-1 (Jan 1, 2004)
http://www.nature.com/embor/index.html
4. Costs and business models in scientific research publishing: A report commissioned by
the Wellcome Trust (April 2004); http://www.wellcome.ac.uk/assets/wtd003184.pdf first citation
in article
5. Optics Express http://www.opticsexpress.org/
6. Publication Business Models and Revenue, in Electronic Scientific, Technical, and
Medical Journal Publishing and Its Implications: Report of a Symposium. The National
Academies Press (2004), ISBN: 0-309-09161-6. http://www.nap.edu/
7. A. Schaffer, Open Access: Should scientific articles be available online and free to the
public? Slate Magazine (Posted Tuesday, Dec. 14, 2004, at 1:21 PM) http://slate.msn.com
8. Budapest Open Access Initiative. http://www.soros.org/openaccess/index.shtml
9. Implementation of Policy on Enhancing Public Access to Archived Publications
Resulting from NIH-Funded Research NIH Notice Number: NOT-OD-05-045 (April 29, 2005);
http://grants.nih.gov/grants/guide/notice-files/
10. Washington Post "President Vows Veto On Stem Cell Research" (May 21, 2005)
http://www.washingtonpost.com/wp-dyn/content/article/2005/05/20/AR2005052000482.html
528
11. New York Academy of Science ebriefings The Coming Revolution in the Publication of
Scientific Papers. (Nov. 25, 2003);
http://www.nyas.org/ebriefreps/main.asp?intSubSectionID=587
12. Public Library of Science. http://www.plos.org/journals/
529
10.28. Scientific citation indices are useful in evaluating medical

physicists for promotion and tenure
David W. O. Rogers and William R. Hendee
OVERVIEW
Citation indices (CIs), which can be obtained online from the Institute for Scientific Information
(ISN) Web of Knowledge by subscription or through your hospital or college library, are
measures of how frequently scientific publications are cited in subsequent articles by other
authors. They can be viewed as an impact factor for an author's publications. In some institutions,
CIs are evaluated by Rank and Tenure Committees that are considering individuals for
promotion. Some well-recognized medical physicists support this practice, while others believe it
puts individuals in highly-specialized disciplines (such as medical physics) at a career
disadvantage. This controversy is the subject of this month's Point/Counterpoint.
Arguing for the Proposition is David W. O. Rogers, Ph.D. Dr. Rogers holds a Canada Research
Chair in Medical Physics in the Physics Department of Carleton University in Ottawa. Previously
he worked at the National Research Council of Canada where he headed the Ionizing Radiation
Standards group from 1985. He obtained his Ph.D. in experimental nuclear structure physics
under A.E. Litherland at the University of Toronto in 1972. His research centers around radiation
dosimetry including clinical dosimetry protocols and the development and application of Monte
Carlo techniques to medical physics problems. He currently serves as Deputy Editor of Medical
Physics.
Arguing against the Proposition is William R. Hendee, Ph.D. Dr. Hendee received the Ph.D.
degree in physics from the University of Texas. He joined the University of Colorado, ultimately
serving as Professor and Chair of Radiology for several years. In 1985 he moved to Chicago as
Vice President of Science and Technology for the American Medical Association. In 1991 he
joined the Medical College of Wisconsin, where he serves as Dean of the Graduate School of
Biomedical Sciences and President of the MCW Research Foundation. His faculty appointments
are Professor and Vice Chair of Radiology, and Professor of Bioethics, Biophysics, and
Radiation Oncology. He also is Professor of Biomedical Engineering at Marquette University and
Adjunct Professor of Electrical Engineering at the University of Wisconsin-Milwaukee.
FOR THE PROPOSITION: David Rogers, Ph.D.
Opening Statement
If properly used, citation analysis can be a useful tool for a committee which is assessing a
medical physicist since it gives one type of indication of the impact of the physicist's research,
which we will assume is part of the physicist's job description.
530
Citation analysis is useful because it provides an assessment of the impact of a researcher by a

broad range of his/or her international and mostly impartial peers, rather than by committee
members who most likely do not understand the research. It is a better indicator of the value of
the research than a publication count since a persistent author can almost always get even a poor
paper published.
Citation analysis can be used well or it can be used badly. One must be vigilant to avoid using the
tool badly. So what are the ground rules for effective use of citation analysis? 1) Citation analysis
must not be the only indicator used. The impact of a piece of work may not be reflected by
citations, such as if a new technique is recommended in a Task Group report which subsequently
receives the majority of the citations. 2) The citation counts must be appropriately compared to
similar counts for a body of peers. Some perspective can be gained looking at the citation counts
for the most cited papers in Medical Physics and PMB which were recently reported by
Patterson.1,2 These are a baseline on the upper limits on citation counts in the field. Even these
most cited papers have relatively low citation counts compared to some other fields. 3) Self-
citations must be removed from the counts. 4) The researcher being evaluated should be asked to
provide a list of sources to be considered, since papers or reports outside ISI's journal database
are only associated with the first author. For example, under my name you will not find any
citations to the EGS4 manual which I co-authored since they are only listed under the name of
my co-author WR Nelson. 5) One must account for the fact that medical physics research often
has a long time constant, unlike some areas of biology where researchers can sometimes react to
another paper's results in a matter of months. One of my papers was cited nearly twice as often
6 to 10 years after publication as in years 1–5. 6) In common with all evaluations of co-
authored publications, the role of a given author in a published work needs to be assessed—was it
a small part or the driving force for the whole project?
But how useful is citation analysis as an indicator? By going to http://scholar.google.com/ and

typing in a name you can get a very quick indication of the impact of someone's research, as long
as you compare the results to those of peers. However, this free site is not as comprehensive as
the more rigorous results found at the subscription ISI Web of Knowledge.
One myth that must be dispensed with is the argument that an incorrect result will generate more
citations than a correct paper. Only errors by highly regarded authors ever get broadly cited for
the errors they contain, while most errors are just ignored.
In summary, citation analysis can provide a useful insight into the impact of an individual's
research output. This must not be the only criterion used by a promotion committee, but it is a
useful indicator when trying to judge the impact of work which is likely to be outside the
committee's immediate fields of expertise.
Rebuttal
While I agree with many points that Dr. Hendee has made, I believe that the constraints I gave on
what is the appropriate use of citation analysis covers many of his objections. So, for example,
when selecting a proper peer group, account must be taken of the popularity of a given area of
research. So it would not be appropriate to compare someone doing research in IMRT to
someone investigating fundamentals of primary standards of air kerma in x-ray beams. On the
other hand, within a hot mainstream field like IMRT there are researchers whose work has more
impact and this is almost universally signaled by a high citation count. At the same time, there
are many IMRT papers with few citations, despite this being an area with many publications. A
531
promotion committee would have some useful information about a candidate if they knew which
group the candidate's papers belonged to.
While I agree that truly major breakthroughs often do not come from the mainstream of research,
I feel that when these major breakthroughs do occur, they will get widely cited. Einstein did not
work in the mainstream in 1905, but even in his day his work was widely cited as evidenced by
the many people we hear about who disagreed with his work. My opponent's quotation from
Smolin, who makes many valid points, is nonetheless just Smolin's opinion and I would suggest
it is not correct in general. Innovative papers are cited widely if the innovation is of any use,
either to our understanding or in practice. We have all seen papers which were very innovative
but of no value since no-one ever used them. We work in applied physics and if something isn't
used, then what is the value? One characteristic of a strong researcher is to work on problems
where the solution will have some impact. Should we reward someone for an innovative solution
to an unimportant problem? An innovative breakthrough, even in a field outside the mainstream,
will be cited frequently. I agree that there is the rare case of something only being found to be
important much later, but the exception proves the rule.
In short, citation analysis, when done properly, allows a committee to evaluate the impact of an
individual's work in a reasonably unbiased way. Citation analysis must never be the sole means
of evaluation, but it can be a useful tool and a valuable component of the assessment.
AGAINST THE PROPOSITION: William Hendee, Ph.D.
Opening Statement
A citation index is a measure of the frequency with which a particular scientific publication is
referenced by other scientists in their own publications in peer-reviewed journals. Publications
that have a high citation index are widely interpreted as having greater impact on the scientific
progress in a field than those that are referenced less frequently. Within a particular discipline,
publication citations suggest that one's scientific work is contributing to a major pathway of
research in the discipline, and that other scientists consider it to be credible and substantive.
Within limits, citation indices are a measure of the importance of one's scientific effort as viewed
by peers, and often they are interpreted in this manner. However, a high citation index may
reveal primarily that one is working in the mainstream of research in a discipline, and that there
are many other scientists working in the mainstream and citing each other's publications. An
individual conducting research in an area where many others are working will have a higher
citation index for publications than will a person in a discipline where fewer scientists are
publishing. This difference in citation index is more a reflection of the number of researchers in
the field than a measure of the relative importance of the publications. Further, a high citation
index may indicate simply that one is in the mainstream of research rather than at the margin or
on an independent pathway that few researchers are following. That is, the citation index is as
much a measure of conformity as it is a measure of importance of one's work—and in many cases
conformity overrides importance.
Major breakthroughs in science typically do not come from scientists working in the mainstream
of research. Usually, they come from individuals of extraordinary creativity and independence
who ask new questions, recognize unexamined assumptions, or extrapolate ideas from one field
to another. As Smolin has described,3 "Many of Einstein's contemporaries testified that he was
532
not unusually talented mathematically. Instead, what enabled him to make such tremendous
advances was a driving need to understand the logic of nature, tied to a breathtaking creativity
and a fierce intellectual independence. But Einstein does not stand alone. One can cite many
examples showing that big advances in physics come when unusually creative and intellectually
independent individuals ask new questions and forge new directions." Smolin goes on to say:3
"People who develop their own ideas have to work harder for each result, because they are
simultaneously developing new ideas and the techniques to explore them. Hence they often
publish fewer papers and their papers are cited less frequently than those that contribute to
something hundreds of people are doing."
The risk in giving substantial weight to citation indices in evaluating scientists for promotion and
tenure is that decisions may favor those working in the mainstream of well-populated fields of
research, and reflect conformity of the research effort rather than original and independent
thinking. The unusually creative and free-thinking scientist would frequently be penalized by the
citation-index criterion, whereas the mediocre scientist pursuing inquiry along a common
pathway with many others would be rewarded. This distinction is in exactly the wrong direction
if truly creative scientists are to be nurtured, and fields such as medical physics are to thrive in
the academic setting.
Rebuttal
In research institutions, several criteria are used to determine an individual's suitability for
promotion and tenure. They include the level of peer-reviewed research support, the individual's
publication record (number of publications and prestige of the journals in which they appear),
and the stature of the individual as a researcher as attested to by highly-regarded peers. In some
institutions, citation indices also are used.
Medical physicists often do not fare well in these analyses of productivity. Unlike other basic
scientists, many medical physicists have heavy clinical workloads that interfere with their
research efforts. Often they are engaged in teaching graduate students, residents and
technologists, which also takes time from research. Medical physics research is frequently
technology-focused rather than disease-focused, which presents challenges when seeking
research support from the National Institutes of Health. And, finally, medical physics is a niche
specialty within biomedicine, so that citation indices are smaller than those for scientists working
in more-populated disciplines without the constraints facing medical physicists.
Even with these handicaps, many medical physicists are highly-productive researchers, educators
and clinical physicists who deserve to be recognized and honored by the promotion and tenure
process. This recognition requires insight by the rank and tenure committee into the profession of
medical physics, and a willingness to judge physicists as individuals and not as cases to be
evaluated against pre-established measures such as citation indices. Further, the committee must
understand that scientific advances usually are made by individuals working at the margins of a
discipline rather than in the mainstream, where the citation indices are invariably greater.
Intelligent decisions about rank and tenure require extraordinary knowledge and judgment about
the worthiness of individuals. They should not be prejudiced by dependence on criteria that more
often reflect conformity within a discipline rather than a presence at the frontiers of knowledge.
REFERENCES
533
1. M. S. Patterson, "Medical physics top ten," Med. Phys. 31, 682 (2004).
2. M. S. Patterson, "The physics in medicine and biology top ten," Phys. Med. Biol. 49, L1–
L4 (2004).
3. L. Smolin, "Why no `New Einstein'?," Phys. Today 56–57 (2005).
534
10.29. Offering the physics exam to residents early in their residency

is helpful to the residents and the radiology department
Krishnadas Banerjee and Gary T. Barnes
OVERVIEW
A few years ago the rules of the American Board of Radiology were changed to permit residents
to take the physics certification examination as early as the second year of their residency. Some
physicists believe this change encourages early learning of the physics of radiology, which is
helpful in understanding clinical radiology. Others believe that residents should be dissuaded
from taking the examination early, because physics is best learned through repetitive courses.
This controversy is the topic of this month's Point/Counterpoint.
Arguing for the proposition is Krishnadas Banerjee, Ph.D. Dr. Banerjee received his M.Sc in
Physics from Calcutta University. He came to the University of Pittsburgh as a Fulbright Scholar
and completed his Ph.D. in Biophysics in 1966. He did two years of postdoctoral fellowship at
Saha Institute of Nuclear Physics in Calcutta. In 1970 he joined St. Francis Medical Center in
Pittsburgh, Pennsylvania as the Director of Medical Physics. He started a master's degree
program in Medical Physics in collaboration with Carnegie Mellon University. Dr. Banerjee has
been teaching radiology, nuclear medicine physics, and radiation biology to residents since 1970.
He has been a written and oral board examiner for the American Board of Radiology for a
number of years.
Arguing against the Proposition is Gary T. Barnes, Ph.D. Dr. Barnes is currently Professor
Emeritus, Physics and Engineering Division, Department of Radiology, University of Alabama at
Birmingham (UAB) and President of X-Ray Imaging Innovations. He received his Ph.D. in
physics from Wayne State University, Detroit, Michigan and medical physics postdoctoral
training at the University of Wisconsin, Madison. He joined the UAB in 1972, and in 1998
started X-ray Imaging Innovations, a company whose mission is to improve medical x-ray
imaging. From 1976 to 1987 he was chief of the Physics Section and from 1987 to 2002 Director
of the Physics and Engineering Division of the Department of Radiology. He continues to be
involved part-time at UAB Chairing the Radioisotope and Radiation Safety Committee and
teaching radiology residents. Dr. Barnes is a past president of the AAPM and its Southeastern
chapter. He is a Fellow of the American College of Radiology and the American Association of
Physicists in Medicine and is a Diplomate of the ABR (Radiological Physics).
FOR THE PROPOSITION: Krishnadas Banerjee, Ph.D.
Opening Statement
Let me preface my discussion with a little history. Until a few years ago a physician candidate
took the ABR clinical and physics written examinations simultaneously at the beginning of the
last year of residency. If the candidate passed both examinations (s)he was allowed to take the
535
oral clinical examination in the next Spring. If (s)he failed the written physics examination and
passed the clinical written exam, (s)he was also examined in physics at the oral board
examination.
In speaking with my colleagues, I have come to realize that although the teaching of clinical
radiology is well prescribed, the teaching of physics is not so organized. In many institutions, the
same physics course is repeated during each year of residency. The ACGME requires 80 hours
of didactic education in physics and radiation biology for radiologists. Usually, these classes are
held once or twice a week either early in the morning or late in the afternoon after the residents
have finished their clinical duties. In one prestigious university, physics teaching is outsourced to
a consultant physicist over a period of three months before the written physics examination.
Although a solid background in physics is essential to understanding the modern technologies of

CT, MRI, fMRI, PET, and SPECT, most residents are not interested in really understanding
physics. Rather, they are interested in passing the physics portion of the written examination.
Residents often try to learn physics by "cramming" just before the examination. Others attend
review courses or purchase board-review textbooks with the hope of finding answers to questions
asked in the examination.
Many students find it difficult to master the huge volume of knowledge covered in the physics
and clinical portions of the written board examination. So when the opportunity was provided in
1999 to take the physics portion of the written examination separately in the second year of
residency, most residents chose to do so, in an attempt to "divide and conquer" the large amount
of subject matter. By passing the physics portion first, they are then able to concentrate on the
clinical exam in the third and fourth years of residency.
From my perspective there are two clear advantages of dividing the written board examination
into two parts. (1) By concentrating on learning physics concepts in the first year of residency, a
solid physics underpinning is achieved that improves the resident's ability to understand and use
sophisticated technologies such as CT and MRI. This underpinning leads to a greater
appreciation for the applications of physics in clinical radiology. After passing the physics
examination in the second year, physicians can turn their attention to topics of clinical radiology
for the remainder of their residency training. (2) Division of the examination is only part of the
solution to improving the quality of graduate medical education in radiology. Program directors
should consider recruiting candidates with solid physics backgrounds, and training programs
should improve the quality of physics courses taught to residents. The latter involves increasing
the time allotted for physics in resident curricula, ensuring that experienced and knowledgeable
faculty are providing the education, and rewarding these faculty for their teaching contributions.
It also means teaching the applied physics of imaging systems later in the residency as physicians
are scheduled through various clinical rotations.
Rebuttal
I have known my "opponent" in this Point/Counterpoint for more than 20 years, and we have
worked together on written and oral examinations for the American Board of Radiology. He is a
stellar teacher, as evidenced by the teaching chair awarded in his name at the University of
Alabama.
My opponent and I agree that having residents take the physics exam early in their period of
training is helpful to the department and to the teaching of clinical radiology. We differ, however,
536
on the issue of whether this is a constructive way to learn physics. Dr. Barnes believes that taking
the exam early is counter-productive, and even suggests that it would be better to drop physics
entirely from residency training than to offer the exam in the second year of training. He believes
that it is preferable to administer the exam in the fourth year of residency.
I personally believe that residents who take their exam in the fourth year of training do not learn
any more physics than those who take the exam in the second year. This is because the residents
want to learn just enough physics to pass the exam, no matter when it is given. They do their
learning just ahead of the exam, so having residents repeat a physics course for 2 or 3 years
doesn't accomplish very much. In addition, the major challenge in teaching physics is to make
the subject interesting and relevant to the practice of radiology, no matter when the teaching
occurs.
Learning physics early in the residency helps residents understand imaging technology and
radiation safety when they encounter such challenges during clinical rotations. They will be
conscious of the need for dose reduction in fluoroscopy, limited exposures in computed
tomography, restriction of the number of films taken during angiography, and application of the
optimum procedures for data acquisition in ultrasound and magnetic resonance imaging.
In addition, the American Board of Radiology does not require residents to take the physics exam
in the second year of residency. They can take it in the third or even the fourth year if they
choose to do so. Further, a residency program can decide when its residents should take the
physics exam. The policy of the Board is permissive, but not specific; residents can take the
physics exam any time after the first year of residency.
If residency program directors make time available for residents to study physics during their
first year of training, and if physicists who teach residents make an effort to portray physics as
interesting and relevant to the clinical practice of radiology, then I am comfortable that offering
the exam in the second year of residency will not handicap the physics learning process in
radiology.
AGAINST THE PROPOSITION: Gary T. Barnes, Ph.D.
Opening Statement
There is no question that offering the physics exam to residents early in their residency is helpful
to the radiology department. It is also helpful to residents in their moving through the ABR exam
process. It is my position that it is not helpful in their learning a useful level of physics. As a
result it compromises radiology and patient care.
Prior to 1999 residents were eligible to take the written physics and clinical exams early in their
fourth year of residency. If the resident successfully passed both, (s)he was then eligible to take
their oral exams at the end of their fourth year. A problem with scheduling the written exams in
the beginning of the fourth year of residency is that residents were spending significant time in
the latter part of their third year and beginning of their fourth year of residency studying for the
written exams. This, coupled with the fact that they spent time studying for their oral boards in
the latter part of the fourth year, resulted in their taking considerable time away from their
clinical support of the department at the time in their residency when they were most useful to the
537
department. This was a major consideration in allowing residents to take their written exams
earlier in their residency.
A secondary consideration is the role of the written exams. They are a high-pass filter for the oral
exams. That is, residents that are unable to pass the written exams would have little or no chance
of passing their oral exams. Furthermore, giving an oral exam to a candidate who fails is tough
on the candidate and tough on the examiner. For these reasons the written exams were instituted
more than thirty years ago. Previously there was only the oral exam which all residents were
eligible for at the end of their residency, and physics was one of the categories in which they
were examined.
Starting in 1999 the current ABR policy went into effect and residents were eligible to take their
written physics exam in the beginning of their second year of residency and their written clinical
exam in the beginning of their third year. They could elect to take either written exam later, but in
order to be eligible to take their oral exam they have to pass both the written physics and clinical
exams.
Prior to 1999, physics education of residents typically involved one or more courses in the first
two years of residency, followed by a board review course prior to their taking the written
physics exam in the beginning of their last year. In their fourth year of training residents have a
mature understanding of radiology. This maturity was evident in the questions they would ask
and was helpful in their understanding of the role physics plays. Also, a cardinal principle of
education is repetition. The process prior to 1999 allowed for repetition. The current process does
not.
Subsequent to the changes that went into effect in 1999, most residents have elected to take their
physics written exam at the beginning of their second year of residency and early in their
training. There are a number of problems with this that limits their education, negatively impacts
radiology, and negatively impacts patient care. First, after one year during which residents are
struggling to master a vast field, their knowledge of radiology is superficial and their ability to
appreciate the role and importance of physics is even more limited. They have not rotated
through a number of areas such as nuclear medicine or angiography, where a knowledge of
patient and personnel radiation levels, and the interplay of radiation levels and image quality, are
important considerations.
Second, their focus at the end of their first and beginning of their second year is on passing the
written physics exam. They do not want to learn anything that is not related to their doing well on
the exam. Unfortunately, the ABR written physics exam is superficial at best. Residents who
have little understanding of imaging physics and its utility in the practice of radiology still pass
the exam. Memorization and little or no understanding of imaging physics principles is all that
are required. The idea of such an individual acting as a radiation safety officer or being licensed
to practice nuclear medicine is absurd. Residents passing the written physics exam at the present
time often have little knowledge of imaging physics principles, patient dosimetry and radiation
safety. Residents who have a very good understanding of physics often do not do as well on the
exam as residents with a more superficial understanding. They do well but often not as well. In
view of these experiences I advise residents not to think too much. This advice has proved useful.
Third, the cardinal principle of education, "repetition", is not utilized. Residents are exposed to
one meaningful physics course—the board review course. On passing the exam they naturally
538
focus on other more immediate aspects of their residency training. Little or no attention is
subsequently paid by most residents to physics.
At one time a knowledge of radiation and imaging physics principles was one factor that
differentiated radiologists from other physicians doing imaging. This is no longer the case. A
non-radiologist can easily take a two or three day short course in an area of imaging and have a
fund of physics knowledge that exceeds that required to be a diplomat of the ABR. In the late
1980s and early 1990s there were a number of patients who suffered severe burns when
undergoing fluoroscopic procedures.1 The majority, if not all, of these cases involved non-
radiologists. If radiologists practicing at the time had the limited physics training of current
residents, they might have contributed to this problem. It is my experience that the level of
physics in the cardiologists' fluoroscopic clinical competence statement 2 is far more detailed and
extensive that the material currently taught to radiology residents.
I have been teaching physics to radiology residents for more than thirty years. In the UAB
Department of Radiology a Distinguished Faculty Teaching Award was initiated by residents in
my name. Until recently I had been teaching concepts and ideas to residents that will be useful in
their future practice. Now I prepare them to take an exam. It is no longer teaching. It is more akin
to training dogs to jump through hoops. The level of radiological physics they learn to pass the
exam is insufficient to be useful in their future practice. I question whether I am making
productive use of my time. Allowing residents to take the physics exam after one year of
training, and when they pass (as almost all of them do) claim that they have an acceptable level
of understanding of radiological physics is not consistent with reality. It would be more honest to
drop physics and the physics exam from residency training altogether, than to continue the
current hypocrisy.
Through their exams and their policies, the ABR dictates residency training, whether the Board
admits it or not. It is my opinion that the ABR made a mistake in allowing residents to take the
written physics exam after only one year of training. Residents do not learn a level of physics
that will be helpful to them in their future practice. The physics written exam needs to be
scheduled later. A practical compromise would be to allow residents to take both the physics and
clinical written exams in the beginning of their third year of training.
Rebuttal
I agree with many of the points that my esteemed colleague makes: 1) the teaching of physics to
radiology residents is not well prescribed; 2) a solid background in physics is essential to
understanding modern imaging technologies; 3) residents often try to learn physics by
"cramming" just before the board examination; 4) a good physics underpinning leads to a greater
appreciation for the applications of physics in clinical radiology; and 5) after passing the physics
exam (early) in the second year residents can turn their attention to clinical radiology for the
remainder of their training.
Points that I dispute are: 1) most residents are not interested in really understanding physics and
only want to pass the exam; and 2) by concentrating on learning physics concepts in the first year
of residency, a solid physics underpinning is achieved that improves the resident's ability to
understand and use sophisticated technologies such as CT and MRI.
Obviously, all residents want to pass the written physics exam. It has been my experience,
however, that most residents also want to understand physics as it relates to obtaining good
539
images, patient dosimetry and radiation safety. It is the physicists' job to communicate and
demonstrate how physics is useful in these regards.
I contend that residents do not achieve a solid physics underpinning with the exam being offered
early in their residency. Offering the exam early is conducive to "cramming" and not to
understanding. The radiological physics learning experience, as with most learning experiences,
benefits by repetition. One course focusing on passing the exam achieves minimal long-term
retention and understanding. Further, once most residents pass the physics exam, the other
aspects of radiology are so demanding that they have little or no time to learn physics. The
physics that radiology residents learn occurs prior to and not after their passing the physics
written exam.
As a result of the ABR physics written exam being offered early in the residency, residents now
learn less physics than radiologists that completed their residency five or more years ago. To
reverse this trend, the physics exam should be offered later. A practical compromise would be to
allow residents to take both the physics and clinical written exams in the beginning of their third
year of training. Residents would have rotated at least once through all clinical rotations and be
in a better position to understand the role that physics plays. It would allow time for both an
introductory physics course and a board review course, and therefore repetition. If a resident fails
either the written physics or the clinical exam, s(he) could retake the exam in the beginning of
the fourth year and, presuming a pass, could to take the oral exam in June of the fourth year. Prior
to 1999 and for more than 25 years, residents took the physics and clinical exams at the same
time. It was not considered to be a problem because everybody was in the same boat.
REFERENCES
1. T. B. Shope, "Radiation-induced skin injuries from fluoroscopy," Radiographics 16, 1195

(1996).
2. J. W. Hirshfeld et al., "ACCF/AHA/HRS/SCAI clinical competence statement on
physician knowledge to optimize patient safety and image quality in fluoroscopically guided
invasive cardiovascular procedures," J. Am. Coll. Cardiol. 44, 2259 (2004).
540
10.30. Medical physicists should seek employment elsewhere when

resources are insufficient for quality patient care
Wlad T. Sobol and Ivan A. Brezovich
OVERVIEW
Quality assurance is one of the major responsibilities assumed by medical physicists (MPs) in
hospital radiology and radiation oncology departments. The work done by MPs to assure quality
of care is usually performed in the ―background‖ in that it is often done outside normal clinical
hours, and is not directly a source of revenue to the hospital. Consequently, hospital
administrators are often unaware of this vital role played by MPs and hence, when faced with the
need to trim budgets, they sometimes see MPs as ―expendable.‖ This is exacerbated by the fact
that few hospital administrators understand what MPs do. Many MPs are thus faced with a
situation where they do not have sufficient staff to maintain a level of quality assurance they
consider necessary to protect patients. Their choices are to work ridiculously long hours, to
provide ―substandard care,‖ or to ―quit.‖ Whether or not it is best for physicists facing this
situation to seek employment elsewhere is the topic of this month's Point/Counterpoint.
Arguing for the Proposition is Dr. Wlad Sobol. Dr. Sobol received his Ph.D. degree from the
Jagiellonian University in Cracow Magna Cum Laude in 1978. In 1986 he emigrated to the
United States and joined the Department of Radiology at Bowman Gray School of Medicine in
Winston-Salem, NC as an Assistant Professor and, in 1991, he moved to the Department of
Radiology at University of Alabama in Birmingham, where he is Professor of Radiology. At
UAB he has been very actively involved in education, and is Director of the Clinical Medical
Physics (Imaging) Residency Program. In 1999 he served as a President of the Southeastern
Chapter of the AAPM. He was a member of the Board of Editors of Medical Physics for several
years, served on or chaired several AAPM committees, and was Co-Director of the 2001 Summer
School. He chaired the MRI Examination Panel of the ABMP that developed the MRI
examination for the Board. Dr. Sobol is board certified by the American Board of Radiology in
Diagnostic Radiological Physics and by the American Board of Medical Physics in Magnetic
Resonance Imaging Physics and is a Fellow of the AAPM.
Arguing against the Proposition is Ivan Brezovich, Ph.D. Dr. Brezovich received his Ph.D. in
Physics from the University of Alabama in 1977 and has since spent his entire career as a
medical physicist at the UAB, initially in the Department of Diagnostic Radiology and, since
1977, in the Department of Radiation Oncology, where he has been a Full Professor since 1988.
He is also a Professor in the Department of Biomedical Engineering. Dr. Brezovich is Chairman
of the ACMP Reimbursements Committee and has served on many AAPM professional
committees since 1991 and is currently a member of the Professional Economics and
Government and Regulatory Affairs Committees, the Emerging Technology Work Group, and
the European Affairs Subcommittee. In 1994 he served as a President of the Southeastern
Chapter of AAPM. Dr. Brezovich is a Fellow of the AAPM, the ACMP, and ACRO, and a
Diplomate of the ABR in both Therapeutic and Diagnostic Radiological Physics.
541
FOR THE PROPOSITION: Wlad T. Sobol, Ph.D.
Opening Statement
For the sake of keeping the arguments in focus, let me state that I am not concerned with extreme
situations caused by economic, socio-political, or environmental factors as well as acts of war or
effects of natural disasters. I will also exclude conditions where remuneration for services
rendered is not the primary motivation for work. Finally, I will exclude global ventures and focus
on domestic (i.e., U.S.) operations.
This leaves us with a typical healthcare business environment where a medical physicist is a
hired employee tasked with performing specific duties for a set remuneration (specified in an
employment contract as salary and benefits package). In this setting, highly skilled professional
employees are constantly facing negotiations covering a broad spectrum of job-related issues.
This phenomenon is very well known and widely referred to as office politics, 1 games,2 power
plays,3 or even war.4 While negotiating skills are extremely valuable to any working
professional, they can be difficult to master,5,6,7 which explains why an entire field of study,
known as negotiation science, has emerged in recent years. For the sake of the current argument,
we shall enumerate the key elements in negotiations, as identified by negotiation science: 8
- target (or target value) refers to the best possible outcome the negotiating side hopes to achieve;
- reservation point (or reservation price) is the minimum the negotiating party will accept before
abandoning the negotiations;
- best alternative to a negotiated agreement (BATNA) represents the best fallback position
available to the negotiator, should the negotiations fail;
- power in negotiation is the degree of leverage that a particular party has during negotiation;
- interests are the ultimate goals of the negotiating parties;
- positions are the particular stances negotiation parties take to arrive at their ultimate goals;
- anchoring refers to the first offer put forth during negotiations.
It is quite obvious that the negotiating process can be extremely complicated and require a great
deal of preparation, skill, and experience.
Thus, it is very difficult to predict an outcome of any negotiation, especially if the interests of the
parties are undefined (we do not know what ―resources‖ mentioned in the proposition represent,
much less what makes them ―insufficient,‖ especially for ―patient care‖). Furthermore, the
reservation points for different individuals facing the same negotiating targets can be vastly
different, making attempts at detailed discussion moot. Facing such a broad range of possible
negotiating scenarios, one can only conclude that the power of quitting is too big to be
relinquished lightly. Having this option available as a possible BATNA greatly increases the
medical physicist's power in any job-related negotiation. That empowerment allows us to select a
BATNA that truly enhances our negotiating positions by allowing us to think and analyze
situations at strategic, rather then tactical, levels. This view is further supported by reports of real
542
world experiences9 which indicate that many successful professionals have the ability to
recognize hopeless situations and simply quit before further struggle makes things worse.
AGAINST THE PROPOSITION: Ivan A. Brezovich, Ph.D.
Opening Statement
The simplistic suggestion ―just leave‖ flies into the face of every sense of human decency and
responsibility, the very bonds that keep our society together. Apart from the typical disruption
following resignations of key personnel, sudden departures of therapy physicists can be
deadly.10,11 Market forces are too slow to safely bring back a stable equilibrium.
Unlike physicians, medical physicists have many skills that are specific to a facility and are not
readily interchangeable. To safely ―orchestrate the entire treatment process,‖ as an ASTRO Board
Chairman summarized our role in radiotherapy, medical physicists must thoroughly understand
every machine and procedure, as well as the workings of the entire system, including human
dynamics. High-tech equipment like accelerators and treatment planning systems has its specific
flaws, and physicists are expected to detect and safely defuse the ensuing death traps.
Manufacturers limit their product warranties to repairing or replacing defective equipment. The
responsibility for clinical applicability and consequences of malfunctions lies squarely on the
shoulders of the physicist. The margin of safety remains very narrow for a long time when an
experienced physicist leaves, even if the replacement is highly qualified.
Furthermore, while physician training is universal, there are no uniform educational standards for
medical physicists. In many states, cancer clinics can hire any self-proclaimed ―experts‖ with no
demonstrable qualifications and put them in charge of their patients. Other states require a
graduate degree and on-the-job training, but no examination. Even a board-certified Qualified
Medical Physicist (QMP) may not have all the knowledge on which a clinic has learned to
depend. Physicists are expected to check dose prescriptions,12 yet the required knowledge is
neither taught in graduate school nor tested during ABR exams. Physicists who possess such
precious expertise are hard to replace.
If physicists cannot provide quality patient care because their facilities lack adequate equipment
and support staff, or expect an unreasonable workload—42% of medical physicists at cancer
centers work more than 50 h/week13—they may be tempted to seek employment elsewhere.
However, before taking such a drastic step, they need to consider potential consequences. An
experienced physicist will flatly refuse to accept such a position. With no qualified takers, the
clinic may hire one of the abundant recent graduates, who has little more than online training and
is desperate for a first job. Lower salary expenditures and additional revenue from salvage
treatments necessitated by ineffective initial treatments may even give the false impression that
the personnel change was beneficial.
Because of the unique features of medical physics, its practitioners have obligations in addition
to those of other medical specialists. If unable to provide quality patient care, they must make
every conceivable effort to resolve the problems and remain on the job. If resignation becomes
inevitable, these physicists must explicitly inform the physicians and top administrators about the
true reasons for their departure, even if such action adversely affects future employment. Simply
walking out on their patients is not an acceptable option for medical physicists.
543
Rebuttal: Wlad T. Sobol, Ph.D.
I have no quarrel with my fellow debater's first argument pointing out that abrupt departures are
likely to be disruptive and thus are to be avoided. Medical physicists are expected to follow
professional etiquette—these professional rules of conduct demand that the exit be managed in
orderly fashion. But, I thought, our arguments were about ―whether‖ (to quit), while this point is
about ―how‖ (to resign).
The second point implies that experienced medical physicists are so special that, without them,
the entire process would collapse and therefore it is their duty to keep on going no matter what.
This assertion flies in the face of experience that is best summarized by the proverb ―graveyards
are full of irreplaceable people.‖ Every one of us will be replaced; the question is when and how.
The third argument states that if an experienced physicist quits, a lesser-caliber replacement is
likely to follow. This scenario is further amplified by pointing out that no ―good‖ professional
would want a ―bad‖ job. I am dumbfounded; as far as I know, this is how the American society
works—―good‖ jobs are offered to top-level candidates; ―bad‖ jobs will attract mediocre
applicants, if any. If the question is whether a competent, experienced professional would want to
stay in a bad job at a facility whose management consistently ignores pleas for improvement,
then this is an issue of motivation for work, not about the ability to resign. In other words, this
argument is about setting a reservation point, not for establishing a BATNA.
In closing, I prefer professional freedom over restrictions. I choose to view my fellow medical
physicists as highly trained, responsible, and ethical professionals. I want them to have a freedom
of choice that includes walking away from a hopeless cause. I trust their judgment and their
ability to make wise decisions.
Rebuttal: Ivan A. Brezovich, Ph.D.
I never said that quitting should not be an option for therapeutic medical physicists. With proper
precautions, departures may even benefit patients, at least in the long run. However, using my
distinguished colleague's terminology, quitting by itself does not confer upon physicists
sufficient ―power in negotiations‖ to achieve this ―reservation point,‖ viz. the minimum amount
of resources for patient care that they can accept.
For diagnostic physicists, whose responsibilities for patients are primarily indirect through
assuring image quality, departures should be less problematic. If you get replaced by a ―cheap‖
physicist, radiologists can recognize the deteriorating performance of equipment before patients
get hurt. Disruptions of patient flow and higher equipment maintenance cost will pressure
administrators to rectify the inadequacies that led to departure.
Therapeutic physicists, on the other hand, are directly responsible for human lives. Except for
extreme cases of overdosing,10,11 a ―cheap‖ replacement may even be a financial asset to the
facility. The fact that one-third of sophisticated radiation treatments may not be delivered as
prescribed14 demonstrates the difficulty to even recognize poor quality in radiotherapy. High-
quality physicists, therefore, have special ethical obligations.
Medical specialists are aware that any dissatisfied, overly fatigued person—about 15% of
physicists spend more than 55 h/week on the job13—is prone to make mistakes. They also know
how faulty, outdated hardware and software impair treatment quality, and that a new physicist
544
will have even greater difficulties dealing with the inadequacies. Quality physicists should
consider quitting only when it is the ―best alternative to a negotiated agreement (BATNA),‖ after
their negotiations to obtain the resources required for quality patient care have failed. To benefit
patients, departures must be accompanied by a detailed explanation for their reasons. Simply
walking off the job is highly irresponsible for a therapeutic physicist.
REFERENCES
1. A. DuBrin, Winning Office Politics (Prentice Hall, Englewood Cliffs, NJ, 1990).
2. C. J. Sindermann, Winning the Games Scientists Play (Plenum, New York, London,
1982).
3. R. Greene, The 48 Laws of Power (Viking Penguin, New York, 1998).
4. R. Greene, The 33 Strategies of War (Viking Penguin, New York, 2007).
5. M. C. Donaldson and M. Donaldson, Negotiating for Dummies (IDG Books, Foster City,
CA, 1996).
6. S. R. Covey, The 7 Habits of Highly Effective People, Restoring the Character Ethics
(Simon & Schuster, New York, 1990).
7. S. Deep and L. Sussman, What to Say to Get What You Want (Addison-Wesley, Reading,
MA, 1992).
8. J. W. Berlin and F. J. Lexa, ―Negotiation techniques for health care professionals,‖ J.
Am. Coll. Radiol. 4, 487–491 (2007).
9. S. Godin, The Dip. A Little Book that Teaches You when to Quit (and when to Stick)
(Penguin Group, New York, 2007).
10. L. Stern Rubin, ―The Riverside radiation tragedy,‖ in Columbus Monthly, April 1978
issue, pp. 52–66.
11. L. C. Cohen, T. E. Schultheiss, and R. C. Kennaugh, ―A radiation overdose incident:
Initial data,‖ Int. J. Radiat. Oncol., Biol., Phys. 33, 217–224 (1995).
12. AAPM Report No. 38, ―Statement on the role of a physicist in radiation oncology,‖
(American Institute of Physics, Inc., New York, 1993), p. 8.
13. AAPM Professional Information Report, Calendar Year 2006, p. 17
[http://www.aapm.org/AAPMUtilities/download.asp?file=AAPM-Salary06.pdf]
14. G. S. Ibbott, A. Molineu, and D. S. Followill, ―Independent evaluations of IMRT through
the use of an anthropomorphic phantom,‖ Technol. Cancer Res. Treat. 5, 1–7 (2006).

Controversies in Medical Physics

Uploaded by

Copyright:

Available Formats

Controversies in Medical Physics

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Controversies in Medical Physics

Uploaded by

Copyright:

Available Formats

Controversies in Medical Physics:

American Association of Physicists in Medicine

American Association of Physicists in Medicine

© Copyright by the American Association of Physicists in Medicine. All

CHAPTER 1: General Radiation Therapy 1

CHAPTER 2 : Highly Conformal Radiotherapy: IMRT, Tomotherapy, Stereotactic

CHAPTER 4: General Imaging: Image Quality, Magnetic Resonance, Ultrasound 170

CHAPTER 5: X-Ray Imaging 199

CHAPTER 6: Mammography 243

CHAPTER 7: Nuclear Medicine 264

CHAPTER 8: Radiation Protection, Standards and Regulations 279

CHAPTER 9: Education 358

CHAPTER 10: Professional Issues 408

10.1. Medical physicists should position themselves as institutional resources in

Howard I. Amols: Memorial Sloan-Kettering Cancer Center, New York

The Point/Counterpoint series of debates in Medical Physics began in March 1998

Colin G. Orton & William R. Hendee

General Radiation Therapy

1.1. Monte Carlo techniques should replace analytical methods for

Radhe Mohan and John Antolak

FOR THE PROPOSITION: Radhe Mohan, Ph.D.

AGAINST THE PROPOSITION: John Antolak, Ph.D.

1.2. Dm rather than Dw should be used in Monte Carlo treatment

H. Helen Liu and Paul Keall

FOR THE PROPOSITION: H. Helen Liu, Ph.D.

AGAINST THE PROPOSITION: Paul Keall, Ph.D.

1. Clinical experience is Dw-based. Since the introduction of computer-based treatment

2. Dosimetry protocols are Dw-based. As stated in the overview, modern dosimetry

On the issue of increased complexity. Converting Dm to Dw does involve additional computations.

1.3. Over the next decade the success of radiation treatment

C. Clifton Ling and X. Allen Li

FOR THE PROPOSITION: C. Clifton Ling, Ph.D.

AGAINST THE PROPOSITION: X. Allen Li, Ph.D.

BBRTP requires patient-specific information on tumor/normal-tissue biology.

1. S. Levegrun, A. Jackson, M. J. Zelefsky, E. S. Venkatraman, M. W. Skwarchuk, W.

1.4. Image-guided radiotherapy is being overvalued as a clinical tool

FOR THE PROPOSITION: Howard I. Amols, Ph.D.

That image-guided radiotherapy (IGRT) will improve radiotherapy is taken as axiomatic.

Lacking data, we've placed high expectations on an expensive, untested technology. As we

AGAINST THE PROPOSITION: David A. Jaffray, Ph.D.

Rebuttal: Howard Amols, Ph.D.

Rebuttal: David Jaffray, Ph.D.

1. C. Ling, E. Yorke, and Z. Fuks, "From IMRT to IGRT: Frontierland or Neverland,"

2. A. A. Martinez et al., "Improvement in dose escalation using the process of adaptive

1.5. The value of PET/CT is being over-sold as a clinical tool in

FOR THE PROPOSITION: Lei Xing, Ph.D.

AGAINST THE PROPOSITION: Barry Wessels, Ph.D.

1. D. W. Townsend, J. P. Carney, J. T. Yap, and N. C. Hall, "PET/CT today and tomorrow,"

6. J. D. Bradley, F. Dehdashti, M. A. Mintun, R. Govindan, K. Trinkaus, B. Siegel, and J. S.

X. Allen Li and Paul J. Keall

FOR THE PROPOSITION: X. Allen Li, Ph.D.

AGAINST THE PROPOSITION: Paul J. Keall, Ph.D.

Fallacy #1: Respiratory gating is not an established technology

Fallacy #2: Not many sites have or use respiratory gating

Rebuttal: X. Allen Li, Ph.D.

Rebuttal: Paul J. Keall, Ph.D.

Following are my responses to the statements of Dr. Li:

Unacceptable risk of pneumothorax for marker implantation: Percutaneous22 and

1. P. J. Keall, G. S. Mageras, J. M. Balter, R. S. Emery, K. M. Forster, S. B. Jiang, J. M.

13. R. Wagman, E. Yorke, P. Giraud, E. Ford, K. Sidhu, G. Mageras, B. Minsky, and K.