International Journal of Sciences:

Basic and Applied Research

(IJSBAR)
ISSN 2307-4531
(Print & Online)

http://gssrr.org/index.php?journal=JournalOfBasicAndApplied

------------------------------------------------------------------------------------------------------------------------------------------

Quality Assurance and Performance Evaluation of an Elekta

Linear Accelerator: Results From Over a Year QA
Experience in Ghana.
George Felix Acquaha*, Bernhard Schiestla,b
a
Sweden Ghana Medical Centre, Radiation Oncology Physics, P.O.Box CT 3286, Cantonments Accra, Ghana.
b
University Hospital of Innsbruck, Department of Therapeutic Radio Oncology, Austria.
a
Email: [email protected]
b
Email: [email protected]

Abstract

This publication describes the first ever comprehensive Quality Assurance program of a linear accelerator to be
developed and performed in Ghana, West Africa. Sweden Ghana Medical Centre has been operating an Elekta
synergy Platform Linear accelerator over the past years. The QA programs, results and performance evaluations are
presented and analyzed over a twenty (20) month operational period. The checks include daily, weekly, monthly,
and yearly quality assurance procedures as well as patient specific quality assurance in accordance with the centre’s
policy. The results of the evaluation show reproducibility in all quality assurance procedures with an average photon
and electron daily radiation output constancy of (-0.8%±1.2) and (-1.38±0.96) of expected values respectively. The
weekly and monthly radiation output constancy checks had an average deviation of 0.53%±0.4 and a maximum
deviation of 1.9% for photons and an average of 1.5%±0.4 and maximum of 2.6% for electrons. Yearly dosimetry
quality assurance was within 0.8% of calibrated values at commissioning whiles safety and mechanical/optical
checks were functional and within set limits respectively.

Keywords: Action level; QA procedure; Quality assurance; Radiation constancy check; Tolerance levels.

------------------------------------------------------------------------
* Corresponding author.
E-mail address: [email protected]

354
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

1. Introduction

Sweden Ghana Medical Centre (SGMC) is the first and only cancer Center in Ghana and among the very few in
West Africa to operate a linear accelerator (dual energy photon: 6 and 15 MV and a range of electron beams: 6, 10
and 15 MeV). The Center covers a full range of cancer specialties and utilizes modern state-of-the- art equipment
and treatment techniques to provide a 3-dimensional (3D) conformal radiation treatment. With no national or
regional published quality assurance (QA) guidelines to follow, it sometimes becomes very difficult adopting an
international protocol that suits our local settings. It should also be noted that it may be unrealistic to carry out some
of the tests described in international protocols [1, 2, 3, 4] at local centres, either because major investments (human
or technical resources) are required, or because the parameter being tested is not available, or not in clinical use. It is
also the responsibility of a qualified medical physicist to apply these recommendations in a suitable manner [5]. The
delivery of accurate prescribed doses to patients is one of the cardinal objectives of radiotherapy treatment. To
achieve this, it is imperative that effective and efficient quality control (QC) procedures are put in place in every
country where radiotherapy machines are operated. The operation of an effective QC checks will ensure that the
machine characteristics do not deviate significantly from their baseline values acquired at the time of acceptance and
commissioning. This document is intended to provide Ghanaian and to a larger extend West African radiotherapy
centers with streamlined Linac procedures for carrying out requisite QA tests to suit our local conditions. Specific
QA protocols have been adapted from the American Association of Physicists in Medicine Task Group (AAPM TG)
number 40, AAPM TG-45, AAPM TG-142, the European Society of Therapeutic Radiology and Oncology
(ESTRO) and Elekta recommended user checks for the Linac at SGMC.

The success of every radiation therapy crucially depends on the accuracy with which the prescribed dose is
delivered to the tumor volume; thus the dose delivered to the patient should be kept as close as possible to the
prescribed dose. AAPM TG-40 recommends that the dose delivered to the patient be within ±5% of the prescribed
dose [1] and ICRU 50 and 62 (-5%, +7%). Quality Assurance describes programs for checking the performance of
radiotherapy equipment and for measuring the characteristics of the output from such equipment. It has a common
basis in that it specifies the method of testing and test equipment, the parameters to be tested and the frequency of
testing, the responsibilities of different members of staff, the baseline values and tolerances for these values, action
levels and documentation guidelines. A clinical linear accelerator must in all circumstances function within the very
narrow tolerances obtained at the time of acceptance testing [6]. It is therefore expected that a QA program designed
specifically for an institution will meet those standards. For radiation oncology, the QA programs are to maintain the
quality of patient care. The American College of Radiology (ACR) recommends a QA committee be formed with
appropriate personnel (e.g., radiation oncologist, physicist, dosimetrist, therapist, engineer and administrator) [7].
Annual report should be prepared from individual QAs to include but not limited to dosimetry accuracy, mechanical
accuracy, safety, imaging and special Procedures. Acceptance Testing Procedure (ATP) Standards are set as
principles in testing the baseline for future dosimetric measurements in beam performance constancy. This verifies
that the equipment is mechanically functional and operates within certain tolerances from absolute specified values.
Three action levels (level1-Inspection, level 2-scheduled and level 3- immediate/stop treatment/corrective actions)

355
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

are set and followed according to their tolerance.

2. Materials and Methods

The proposed and adopted QA procedure included daily, weekly, monthly, and yearly checks, as well as individual
treatment verifications. The daily QA includes delivery of 500 MU per photon energy as morning Linac warm-up
and subsequent recording of photon and electron dosimetric values of the central axis at SSD of 100cm. The weekly
QA includes measuring photon and electron dosimetric outputs and the wedge factors for photons. The monthly QA
involves a complete and extensive measurement in terms of dosimetry, mechanical/optical and safety aspects. The
annual QA includes a larger sampling of monthly QAs, beam characteristics, planned maintenance (PM), and
analysis of all saved QA worksheet files. The individual treatment verification or “per patient” QA includes
independent MU Calculations and a final check of all plans after the first treatment fraction according to the data in
the record and verification system of the Linac. The procedures used are tailored to meet the needs of the department
using published reports as a guide. All the QA tasks performed with their tolerances are summarized in Table 1.
Based on all these checks, action levels are setted for all radiation constancy and mechanical checks.

2.1. Daily QA

The Linac output is checked on a daily basis prior to the first patient being treated. This test can be broadly
classified into three groups: beam alignment, dosimetry and safety. A qualified medical physicist or radiotherapist
usually performs these daily tests. Elekta recommends [8] that the Linac should be warmed up prior to daily use. A
deviation of less than 3% is desirable for the dosimetric outputs using appropriate diodes. If any of these parameters
are out of tolerance, they are reported to the medical physicists and the Linac clinical operation is suspended until
the problem has been investigated further. With markings on the horizontal and vertical walls, it is ensured that the
sagittal and lateral lasers align with the centre of the marks. The tolerances are 2 mm (that is ±1 mm) for both sets of
lasers as well as the check for central axis rotation with collimator. The optical distance indicator (ODI) is also
check at SSD with a tolerance of 2 mm. The door interlocks and audiovisual monitors are also part of the daily QA.
The results of the daily check must all be within tolerances before the Linac is passed for clinical use.

2.2. Weekly QA

The dosimetric tests are performed with calibrated dosimetry equipment, a phantom and an ionization chamber.
These tests are performed with a high level of accuracy to ensure that small variations are detected. The weekly test
begins with the beam and mechanical alignment test, because deviations here can influence the dosimetry outcome.
The output of the Linac is checked using a calibrated PTW Farmer chamber to ensure that 1 cGy/MU is delivered to
the isocentre under specific reference conditions. A dose of 100 MU is delivered three times and the dosimeter
readings are recorded using a PTW UNIDOS webline electrometer. The output in nC/MU is calculated as follows:

356
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

Output (nC / MU ) = (mean Re ading (nC ) × K TP × Cal.F × K pol × K ele × K s × PCF ) / MU

(1)

Where Cal.F is the calibration factor, (Kpol, Kele and Ks) are the collection efficiency factors and mean Reading is the
average of the electrometer readings. The phantom correction factor (PCF) is taken as 1.0 for a water equivalent
phantom. KTP is the temperature and pressure correction factor. Four different gantry angles (0°, 90°, 180°, and
270°) are chosen for each week’s dosimetry measurements and it is important that the output is within ±3% of the
reference dose. The wedge factor values are also checked for all photon beams using the same reference conditions.

2.3. Monthly QA

Monthly checks involve more extensive checks of the radiation, safety and mechanical parameters. The QA
procedure includes checks on the mechanical system, the image-forming and image detection system, couch
accuracy, accuracy of gantry movement, accuracy of lasers, accuracy of cross wires and collimators, light / radiation
field coincidence, door interlocks, anti-collision devices, optical distance indicator accuracy and emergency stop
buttons. Table 1. describes an overview of the monthly QC checks performed in accordance with these local QA
procedures.

2.4. Annual QA

Annual QA checks are a scaled down version of the commissioning checks. It is a major QC exercise and is
intended to validate the treatment unit for another twelve months. It involves the use of water phantoms to measure
beam profiles and depth dose curves. In addition to yearly recommended QA checks, Planned maintenance checks
[8] are done to keep the system at correct operating conditions and to make sure that the equipment continues to
operate for as long as possible without unplanned corrective maintenance. Although clinical users do not do PM, it
is ensured that those PM programs are fully up to date (6 and 12 monthly PM, 2, 3, 5 and 7 yearly PM performed by
Elekta engineers and technicians).

2.5. Individual Patient QA

Patient QA is an essential part of general QA implementation. Plan printouts should be approved, dated and saved
by the responsible radiation oncologist, with dose prescription clearly written on the plan. This should also be
reviewed and signed by a medical physicist prior to first treatment. To reduce errors on the first setup, an
independent verification of the treatment parameters are made after the first fraction. These include MU check, field
size check, gantry and collimator check, In vivo dosimetry check, port film check etc. All machine parameters used
for patient setup should be correct. In addition, independent calculation of the dose at isocenter and Dmax values on
the central axis for every field is made. When deviations are above the threshold, certain actions are taken. Such as,
changing the isocenteric depth to radiological equivalent depth in inhomogeneous medium, repositioning of the

357
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

Dmax point correctly and correct estimate of the percentages of the field size covered with multi leaf collimators
(MLCs).

Table 1. QA tasks with tolerances and average deviations from analysis.

FREQUENCY PROCEDURE TOLERANCE AVERAGE DEVIATION

Daily Dosimetry
Photon and electron output constancy ±3% -0.8% and -1.38%
Mechanical
Localizing lasers 2mm/±1mm 0.2mm
Collimator rotation with central axis 2mm 0.1mm
Optical distance indicator(ODI) 2mm 0.1mm
Safety
Door interlock Functional Functional
Audiovisual monitors Functional Functional
Weekly Dosimetry
Photon and electron output constancy* ±3% 0.53% and 1.50%
Wedge factor constancy 3% 0.3%
Mechanical
Gantry angle isocenter 2mm 0.1mm
Collimator isocenter 2mm 0.1mm
Beam/field size alignment 2mm 0.2mm
Monthly Dosimetry
Photon and electron output constancy* ±3% 0.45% and 1.52%
Wedge factor constancy 3% 0.3%
Photon and electron beams flatness constancy 3% 1.0% and 0.5%
Photon and electron beams symmetry 3% 1.5% and 1.3%
Calibration of daily Output dosimeter 3% 0.8% and 1.6%
Mechanical
Gantry/collimator angle readouts 1 degree 0.2 degree
Optical distance indicator 2 mm 0.1mm
Field size indicators 2 mm 0.1mm
Treatment couch position indicators 2mm/1 degree 0.1mm and 0 degree
Localizing lasers 2 mm 0.15mm
Light and treatment field coincidence 2 mm 0.7mm
Light field congruence with Collimator 2 mm 0.3mm
Imaging system QA Optimum quality Optimum quality
Safety
Emergency off switches Functional Functional
Door Interlocks Functional Functional
Anti-collision devices Functional Functional
Planned maintenance procedures** Up to date Up to date
Yearly Dosimetry
Photon flatness and symmetry 2% 1.0% and 1.3%
Electron flatness and symmetry 2% 0.4% and 1.1%
Photon/electron output calibration 2% 0.8%
Mechanical
Gantry and treatment table isocenter 2mm 0.5mm
Planned maintenance procedures** Up to date Up to date
Safety
Follow manufacturer’s test procedures** Functional Functional
* With different gantry angles ** Elekta planned maintenance checklist

358
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

3. Results

These results represent the dosimetric, mechanical and safety tests done on the Elekta linear accelerator (photons
and electrons) at SGMC as well as patient specific QAs. Table 1. shows the average deviations from the tolerance
values adopted for the QA procedures performed over the 20-month period. As seen from the tables, the daily
constancy check shows reproducibility and a stable delivery of dosimetric quantity of better than ±3%. For daily
QA, two action levels were established based on the tolerances. Action level 1 was set for deviations of ±3%, where
treatment continues but the senior physicist is notified immediately for inspection. For deviations more than ±3%,
action level 2 was set; where treatment was to be stopped immediately and the problem investigated by the
responsible senior physicist. The results of the daily check were always within tolerances before the Linac was
passed for clinical use.

Table 2. Dosimetric deviations of QA task.

` Wedge factor Flatness Symmetry Dosimeter

Dosimetry Avg Max Avg Max Avg Max Avg Max Avg Max
(%) (%) (%) (%) (%) (%) (%) (%) (%) (%)
Monthly
Photon 0.45 1.30 1.32 1.8 1.04 1.35 1.52 2.31 0.79 1.88
Electron 1.52 2.00 N/A 0.50 0.78 1.34 1.92 1.63 2.60
Weekly
Photon 0.53 1.90 0.35 1.85 N/A N/A N/A
Electron 1.50 2.60 N/A
Daily
Photon -0.80 -3.00 N/A N/A N/A N/A
Electron -1.38 -4.00

4. Discussion

For daily QA tests, those parameters which could seriously affect patient positioning and therefore the registration
of the radiation field and target volume (lasers, ODI); patient dose (output constancy) and safety (door interlock and
audiovisual contact) were included. From table 1, all these parameters checked were within the acceptable limits and
in good working condition. The radiation dosimetry checks are performed with calibrated diodes; hence higher
deviations expected. Weekly dosimetric QA checks were very similar to the daily but performed with a sensitive
ionization chamber to confirm the daily radiation outputs. As seen in the figure 1, the variations of weekly photon
dosimetric QA values over time were within 2% of the ionisation chamber-measured reference value, although there
appears to be a slight positive trend over time for both energies. With over 98 percent of our patients been treated
with photons, it is a clear evidence of the fact that there is consistency of dose delivered to patients over the

359
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

evaluation period. Monthly photon dosimetric QA values also had similar variations and were within 2% of
reference value. As expected all weekly mechanical checks were all within tolerance.

All the tests included in the tables are important for ensuring that the equipment is suitable for high quality and safe
radiation treatments. As seen in the table 2, the variations of the weekly dosimetric values over time were within
2.6% of the ionisation chamber-measured reference values, with a slightly upward trend for electrons. Variation of
gantry angles did not have any dosimetric effect on the outcome of these QA checks. A wedge factor constancy of
0.35% average was recorded for the photon energies. Collimator and gantry isocenters which could affect dosimetric
outcome both had an average deviation of 0.1mm from reference values. For monthly we include more refined
testing of parameters which will either have a smaller impact on the patient (e.g., treatment couch indicators) or
have lower likelihood of changing over a month (e.g., light and radiation field or beam flatness). Monthly
measurements over the period show an average percentage difference of 0.45% for the photons and 1.52% for the
electron beam as compared to the values obtained at commissioning. The maximum difference between dosimetric
QA values and reference values is about 1.3% for photon beam and 2% for the electron beam. Values which do not
follow the normal or expected trends were repeated under the same conditions. The differences as expected were
mainly due to procedural errors or wrong input data. For the yearly QA checks, one of the dosimetric parameters
that required monitoring was the beam output, specified as the absorbed dose on the central axis under reference
conditions. This was to validate the absolute dose calibrations of megavoltage photon and electron beams using the
International Atomic Energy Agency (IAEA) technical report series (TRS 398) [9]. The deviation was below
tolerance value with an average deviation of 0.8% from commissioning values. Using IBA OmniPro-ImRT software
and the MatriXX, the flatness and symmetry of the various beams and energies were obtained using the IEC
protocols. From Fig.2; (a) is the profile for 6MV photon, (b) for 15MV photon, (c) for 6MeV electron, (d) for
10MeV electron and (d) for 15MeV electron beam. All the monthly and yearly beam flatness and symmetry were
within set tolerance values. The less sensitive diodes are used for quick daily checks and were operating within
prescribed tolerances of 3%. Calibration of daily Output dosimeter was done by comparing dosimeter values with
the equivalent reading obtained from the calibrated ionisation Chamber. Average deviations of 0.8% and 1.6% were
obtained for photon and electron dosimeter calibration factors respectively. If the deviation is less than ±3%, no
action is taken, if the deviation is higher than ±3%, the dose meter is recalibrated.

A custom built LAS VAGAS QA Phantom was used in analyzing the image forming and detection system. Digital
images were obtained and the number of holes at every row and columns apart were evaluated using the iView
software tool. Monthly analysis of all the images produced an optimum image quality. The images were also
observed for distortions.

QA for individual Plans had deviations below the defined tolerance value with an average dose deviation at
isocentre of 2.4% and 1.7% for Doses at maximum depths (Dmax) using an independent MU program. Other
independent verification of the treatment parameters made before and after the first fraction were within allowable

360
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

limits. The experimental techniques for performing these QA tests are not discussed in this work, as they are
described in a number of publications cited.

Fig. 1. Variation of weekly photon dosimetry over the 20-month period for 6 MV and 15MV

Fig.2. Sample beam profiles for analyzing symmetry and flatness.

361
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

5. Conclusion

The QA procedures were structured to meet the clinical requirement for accuracy; necessary to achieve optimum
treatment. This quality requires a multidisciplinary team effort, as it is concerned with the reduction of errors and
uncertainties in every aspect of the radiotherapy process. From this stems the need to perform stringent and regular
QC checks, in terms of dose accuracy, for the Linac to deliver specific doses to the tumor. For the preparation of this
manual, it was decided to rely as much as possible on well-known published national and international guidelines [1,
2, 3, 4] as well as the Elekta recommended checks [8]. The structure of this document was intended to be as general
as possible (logistically and human resources wise), in order to be easily adapted to the different structures of
Ghanaian and West African radiotherapy departments. At the same time, this document aimed to introduce a
reasonable level of uniformity for Linac QA methods throughout Ghana and West Africa. It should be noted that,
these procedures, however, be considered as recommendations and not as mandatory Checks. The daily, weekly and
monthly tests discussed here should be adequate to detect any potential problems in the operation of the Linac and
delivering of accurate prescribed doses. Reproducibility of these procedures over the 20-month period for the Linac,
as well as the accuracy and repeatability of the treatment plans generation and dose delivery were excellent.

After analyzing and evaluating the results of all the QA checks done over the 20 month period, it became prudent to
introduce additional QA checks to the existing document. SGMC radiotherapy policies and procedure manuals were
also to be reviewed yearly and updated as QA procedures change. An external quality audit team (probably, ESTRO
equal or the local radiation protection institute (RPI)) is to investigate our local QA protocols and advice
accordingly. Adherence to the QA outlined is to be followed unless there is demonstrable reason to modify them.
Parameters which show large deviations from their baseline values should be given special attention and checked
more frequently. Alternatively, if careful and extended monitoring demonstrates that a parameter does not change,
or hardly changes at all, then the frequency for monitoring this parameter could be reduced. Although it is difficult
to recommend how long a parameter should be monitored before decreasing the test frequency.

It was recommended that the magnitude of any deviations from the reference value were to be recorded for each test,
rather than simply using a tick/OK to confirm that the test has been carried out and the results lie within the allowed
tolerance limits. This enables trends to be seen and actions taken before the tolerances are exceeded. Finally, the
execution of the checks requires the appropriate allocation of time and of human resources, which directly affects
the daily workload of the treatment machines. Therefore, the quality assurance procedures should be considered as
an integral part of the machine workload and the required time should be allocated within the normal working hours.

Acknowledgements

This manuscript is based on work done at the Sweden Ghana Medical Centre, Accra, Ghana from February 2012 to
October 2013. I wish to express my gratitude to the oncologists, physicists, dosimetrist, therapists, radiographers,
engineer, and other members of the treatment team and all staff of SGMC. This work is also dedicated to the

362
 International Journal of Sciences: Basic and Applied Research (IJSBAR) (2014) Volume 14, No 1, pp 354-363

memory of the late Mr. Emmanuel Kwaku Nani (a pioneer of Medical Physics in Ghana, my mentor and a great
friend).

References

[1] GJ Kutcher, L Coia, M Gillin, et al, “Comprehensive QA for radiation oncology: Report of AAPM Radiation
Therapy Committee Task Group 40”, Med Phys. vol.21, 1994, pp.581–618.

[2] EE Klein, J Hanley, J Bayouth, et al, “Task group 142 report: quality assurance of medical accelerators”, Med
Phys. Vol. 36, 2009, pp. 4197-4212.

[3] R Nath, PJ Biggs, FJ Bova, et al, “AAPM code of practice for radiotherapy accelerators: Report of AAPM
Radiation Therapy Task Group 45”, Med Phys. Vol. 21, 1994, pp. 1093-1121.

[4] Aletti P et al., Recommendations for a quality assurance programme in external radiotherapy, ESTRO,
Booklet No. 2, 1995.

[5] Swiss Society of Radiobiology and Medical Physics. Quality control of medical electron Accelerators,
recommendation No. 11, 2003, pp. 1-31.

[6] BG Haffty and LD Wilson. Handbook of radiation oncology: basic principles and clinical protocols. Jones &
Bartlett Learning, 2008.

[7] Khan FM. The physics of radiation therapy 4th ed. Lippincott Williams & Wilkins, Philadelphia, 2010, pp.
375-404.

[8] Elekta Updated Planned Maintenance Procedures, User notice No. U274, 2011.

[9] International Atomic Energy Agency (IAEA), absorbed dose determination in external beam radiotherapy.
Code of practice for high energy photon beams, TRS-398. Vienna, Austria, IAEA; 2000.

363