Advanced Neuro MR Techniques and Applications 1st Edition All Chapter Instant Download

Full download test bank at ebookmeta.
com
Advanced Neuro MR Techniques and Applications 1st

Edition
For dowload this book click LINK or Button below
https://ebookmeta.com/product/advanced-neuro-mr-
techniques-and-applications-1st-edition/
OR CLICK BUTTON
DOWLOAD EBOOK
Download More ebooks from https://ebookmeta.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Advanced AI Techniques and Applications in

Bioinformatics 1st Edition Loveleen Gaur
https://ebookmeta.com/product/advanced-ai-techniques-and-
applications-in-bioinformatics-1st-edition-loveleen-gaur/
Advanced Materials Polymers and Composites New Research

on Properties Techniques and Applications 1st Edition
Omari V. Mukbaniani (Editor)
https://ebookmeta.com/product/advanced-materials-polymers-and-
composites-new-research-on-properties-techniques-and-
applications-1st-edition-omari-v-mukbaniani-editor/
Advanced Painting Techniques 1st Edition Ruben Gonzalez
https://ebookmeta.com/product/advanced-painting-techniques-1st-
edition-ruben-gonzalez/
Advanced Control of Power Converters: Techniques and

Matlab/Simulink Implementation (IEEE Press Series on
Control Systems Theory and Applications) 1st Edition
Hasan Komurcugil
https://ebookmeta.com/product/advanced-control-of-power-
converters-techniques-and-matlab-simulink-implementation-ieee-
press-series-on-control-systems-theory-and-applications-1st-
Artificial Intelligence and Industrial Applications:
Algorithms, Techniques, and Engineering Applications
1st Edition Tawfik Masrour
https://ebookmeta.com/product/artificial-intelligence-and-
industrial-applications-algorithms-techniques-and-engineering-
applications-1st-edition-tawfik-masrour/
Advanced Integration Techniques Third Version Edition

Zaid Alyafeai
https://ebookmeta.com/product/advanced-integration-techniques-
third-version-edition-zaid-alyafeai/
Features and Management of Acute and Chronic Neuro

Covid Cascella
https://ebookmeta.com/product/features-and-management-of-acute-
and-chronic-neuro-covid-cascella/
Basic and Advanced Laboratory Techniques in

Histopathology and Cytology 2nd Edition Pranab Dey
https://ebookmeta.com/product/basic-and-advanced-laboratory-
techniques-in-histopathology-and-cytology-2nd-edition-pranab-dey/
Data Science: Techniques and Intelligent Applications

1st Edition Pallavi Vijay Chavan
https://ebookmeta.com/product/data-science-techniques-and-
intelligent-applications-1st-edition-pallavi-vijay-chavan/
Advanced Neuro MR
Techniques and
Applications
Advances in Magnetic Resonance
Technology and Applications Series
Series Editors
In-Young Choi, PhD

Department of Neurology, Department of Radiology, Department of Molecular & Integrative
Physiology, Hoglund Biomedical Imaging Center, University of Kansas Medical Center,
Kansas City, KS, United States
Peter Jezzard, PhD

Wellcome Centre for Integrative Neuroimaging, Nuffield Department of Clinical Neurosciences
University of Oxford, Oxford, United Kingdom
Brian Hargreaves, PhD

Department of Radiology, Department of Electrical Engineering, Department of Bioengineering
Stanford University, Stanford, CA, United States
Greg Zaharchuk, MD, PhD

Department of Radiology, Stanford University, Stanford, CA, United States
Titles published:
Volume 1 – Quantitative Magnetic Resonance Imaging – Edited by Nicole Seiberlich and Vikas Gulani
Volume 2 – Handbook of Pediatric Brain Imaging – Edited by Hao Huang and Timothy Roberts
Volume 3 – Hyperpolarized Carbon-13 Magnetic Resonance Imaging and Spectroscopy – Edited by

Peder Larson
Volume 4 – Advanced Neuro MR Techniques and Applications – Edited by In-Young Choi and Peter
Jezzard
Visit the Series webpage at https://www.elsevier.com/books/book-series/advances-in-magnetic-

resonance-technology-and-applications
Advanced Neuro MR
Techniques and
Applications
Edited by
In-Young Choi
Department of Neurology
Hoglund Biomedical Imaging Center
University of Kansas Medical Center
Kansas City, KS, United States
Peter Jezzard
Wellcome Centre for Integrative Neuroimaging
University of Oxford
Oxford, United Kingdom
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1650, San Diego, CA 92101, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
Copyright © 2021 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopying, recording, or any information storage and retrieval system, without permission in writing from the
publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our
arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found
at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may
be noted herein).
Notices
Knowledge and best practice in this ﬁeld are constantly changing. As new research and experience broaden our
understanding, changes in research methods, professional practices, or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any
information, methods, compounds, or experiments described herein. In using such information or methods they should be
mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any
injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or
operation of any methods, products, instructions, or ideas contained in the material herein.
Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library
ISBN: 978-0-12-822479-3
For information on all Academic Press publications

visit our website at https://www.elsevier.com/books-and-journals
Publisher: Mara Conner

Acquisitions Editor: Tim Pitts
Editorial Project Manager: Emily Thomson
Production Project Manager: Prem Kumar Kaliamoorthi
Designer: Greg Harris
Typeset by VTeX
Contents
List of contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
PART 1 FAST AND ROBUST IMAGING

CHAPTER 1 Recommendations for neuro MRI acquisition strategies . . . . . . . . . . . . . . . . 3
Tony Stöcker
CHAPTER 2 Advanced reconstruction methods for fast MRI . . . . . . . . . . . . . . . . . . . . . . . 21
Florian Knoll
CHAPTER 3 Simultaneous multi-slice MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Andreia S. Gaspar, Ana R. Fouto, and Rita G. Nunes
CHAPTER 4 Motion artifacts and correction in neuro MRI . . . . . . . . . . . . . . . . . . . . . . . . 53
André van der Kouwe
PART 2 CLASSICAL AND DEEP LEARNING APPROACHES TO NEURO

IMAGE ANALYSIS
CHAPTER 5 Statistical approaches to neuroimaging analysis . . . . . . . . . . . . . . . . . . . . . 71
Jeanette A. Mumford
CHAPTER 6 Image registration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Hui Zhang
CHAPTER 7 Image segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Carole Helene Sudre
PART 3 DIFFUSION MRI

CHAPTER 8 Diffusion MRI acquisition and reconstruction . . . . . . . . . . . . . . . . . . . . . . . . 109
Hua Guo
CHAPTER 9 Diffusion MRI artifact correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Jesper L.R. Andersson
CHAPTER 10 Diffusion MRI analysis methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Szabolcs David, Joost Verhoeff, and Alexander Leemans
CHAPTER 11 Diffusion as a probe of tissue microstructure . . . . . . . . . . . . . . . . . . . . . . . . 157
Yaniv Assaf and Daniel Barazany
PART 4 PERFUSION MRI

CHAPTER 12 Non-contrast agent perfusion MRI methods . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Matthias Günther
v
vi Contents
CHAPTER 13 Contrast agent-based perfusion MRI methods . . . . . . . . . . . . . . . . . . . . . . . . 195

Laura C. Bell, Sudarshan Ragunathan, and Anahita Fathi Kazerooni
CHAPTER 14 Perfusion MRI: clinical perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Kevin Yuqi Wang, Eric K. van Staalduinen, and Greg Zaharchuk
PART 5 FUNCTIONAL MRI

CHAPTER 15 Functional MRI principles and acquisition strategies . . . . . . . . . . . . . . . . . . 231
Martina F. Callaghan and Nadège Corbin
CHAPTER 16 Functional MRI analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Wei Zhang and Janine Bijsterbosch
CHAPTER 17 Neuroscience applications of functional MRI . . . . . . . . . . . . . . . . . . . . . . . . 261
Essa Yacoub and Luca Vizioli
CHAPTER 18 Clinical applications of functional MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Natalie L. Voets
PART 6 THE BRAIN CONNECTOME

CHAPTER 19 The diffusion MRI connectome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Michiel Cottaar and Matteo Bastiani
CHAPTER 20 Functional MRI connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Eugene Duff
CHAPTER 21 Applications of MRI connectomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Jessica S. Damoiseaux, Andre Altmann, Jonas Richiardi, and
Sepideh Sadaghiani
PART 7 SUSCEPTIBILITY MRI

CHAPTER 22 Principles of susceptibility-weighted MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
Sagar Buch, Yongsheng Chen, and E. Mark Haacke
CHAPTER 23 Applications of susceptibility-weighted imaging and mapping . . . . . . . . . . . 359
Stefan Ropele
PART 8 MAGNETIZATION TRANSFER APPROACHES

CHAPTER 24 Magnetization transfer contrast MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
Seth A. Smith and Kristin P. O’Grady
CHAPTER 25 Chemical exchange saturation transfer (CEST) MRI as a tunable relaxation
phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Moritz Zaiss, Felix Glang, and Kai Herz
CHAPTER 26 Clinical application of magnetization transfer imaging . . . . . . . . . . . . . . . . . 403
Francesca Bagnato
Contents vii
PART 9 QUANTITATIVE RELAXOMETRY AND PARAMETER

MAPPING
CHAPTER 27 Quantitative relaxometry mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Mark D. Does
CHAPTER 28 MR fingerprinting: concepts, implementation and applications . . . . . . . . . . . 435
Dan Ma
CHAPTER 29 Quantitative multi-parametric MRI measurements . . . . . . . . . . . . . . . . . . . . . 451
Gunther Helms
PART 10 NEUROVASCULAR IMAGING

CHAPTER 30 Neurovascular magnetic resonance angiography . . . . . . . . . . . . . . . . . . . . . 469
Kevin M. Johnson
CHAPTER 31 Neurovascular vessel wall imaging: new techniques and clinical applications 485
Chun Yuan, Mahmud Mossa-Basha, Zachary Miller, and Zechen Zhou
PART 11 ADVANCED MAGNETIC RESONANCE SPECTROSCOPY

CHAPTER 32 Single voxel magnetic resonance spectroscopy: principles and applications 503
Ivan Tkáč and Gülin Öz
CHAPTER 33 Magnetic resonance spectroscopic imaging: principles and applications . . . 519
Wolfgang Bogner, Bernhard Strasser, Petr Bednarik, Eva Heckova,
Lukas Hingerl, and Gilbert Hangel
CHAPTER 34 Non-Fourier-based magnetic resonance spectroscopy . . . . . . . . . . . . . . . . . 537
Peter Adany, In-Young Choi, and Phil Lee
PART 12 ULTRA-HIGH FIELD NEURO MR TECHNIQUES

CHAPTER 35 Benefits, challenges, and applications of ultra-high field magnetic resonance 553
Karin Markenroth Bloch and Benedikt A. Poser
CHAPTER 36 Neuroscience applications of ultra-high-field magnetic resonance imaging:
mesoscale functional imaging of the human brain . . . . . . . . . . . . . . . . . . . . 573
Jonathan R. Polimeni
CHAPTER 37 Clinical applications of high field magnetic resonance . . . . . . . . . . . . . . . . . 589
Louise Ebersberger, Mark E. Ladd, and Daniel Paech
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
This page intentionally left blank
List of contributors
Peter Adany
Hoglund Biomedical Imaging Center, University of Kansas Medical Center, Kansas City, KS,
United States
Andre Altmann
Centre for Medical Image Computing (CMIC), Department of Medical Physics and Biomedical
Engineering, University College London (UCL), London, United Kingdom
Jesper L.R. Andersson
FMRIB Centre, University of Oxford, Oxford, United Kingdom
Yaniv Assaf
School of Neurobiology, Biochemistry and Biophysics, Faculty of Life Science, Tel Aviv University,
Tel Aviv, Israel
Sagol School of Neuroscience, Tel Aviv University, Tel Aviv, Israel
The Strauss Center for Computational Neuroimaging, Tel Aviv University, Tel Aviv, Israel
Francesca Bagnato
Neuroimaging Unit, Neuroimmunology Division, Neurology Department, Vanderbilt University
Medical Center, Nashville, TN, United States
Department of Neurology, Nashville VA Medical Center, TN Valley Healthcare System, Nashville,
TN, United States
Daniel Barazany
The Strauss Center for Computational Neuroimaging, Tel Aviv University, Tel Aviv, Israel
Matteo Bastiani
Wellcome Centre for Integrative Neuroimaging (WIN), Oxford Centre for Functional MRI of the
Brain (FMRIB), University of Oxford, Oxford, United Kingdom
Sir Peter Mansfield Imaging Centre (SPMIC), University of Nottingham, Nottingham,
United Kingdom
Petr Bednarik
High-field MR Center, Department of Biomedical Imaging and Image-guided Therapy, Medical
University of Vienna, Vienna, Austria
Laura C. Bell
Division of Neuroimaging Research, Barrow Neurological Institute, Phoenix, AZ, United States
Janine Bijsterbosch
Department of Radiology, Washington University School of Medicine, Washington University in
St. Louis, St. Louis, MO, United States
ix
x List of contributors
Wolfgang Bogner
Sagar Buch
Department of Radiology, Wayne State University School of Medicine, Detroit, MI, United States
Martina F. Callaghan
Wellcome Centre for Human Neuroimaging, UCL Queen Square Institute of Neurology, UCL,
London, United Kingdom
Yongsheng Chen
Department of Neurology, Wayne State University School of Medicine, Detroit, MI, United States
In-Young Choi
United States
Department of Neurology, University of Kansas Medical Center, Kansas City, KS, United States
Nadège Corbin
Wellcome Centre for Human Neuroimaging, UCL Queen Square Institute of Neurology, UCL,
London, United Kingdom
Centre de Résonance Magnétique des Systèmes Biologiques, UMR5536, CNRS/University
Bordeaux, Bordeaux, France
Michiel Cottaar
Wellcome Centre for Integrative Neuroimaging (WIN), Oxford Centre for Functional MRI of the
Brain (FMRIB), University of Oxford, Oxford, United Kingdom
Jessica S. Damoiseaux
Institute of Gerontology, Department of Psychology, Wayne State University, Detroit, MI, United
States
Szabolcs David
PROVIDI Lab, Image Sciences Institute, University Medical Centre Utrecht, Utrecht,
The Netherlands
Department of Radiation Oncology, University Medical Centre Utrecht, Utrecht, The Netherlands
Mark D. Does
Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States
Institute of Imaging Science, Vanderbilt University Medical Center, Nashville, TN, United States
Eugene Duff
Department of Paediatrics, and Wellcome Centre for Integrative Neuroimaging, University of
Oxford, Oxford, United Kingdom
List of contributors xi
Louise Ebersberger
German Cancer Research Center (DKFZ), Division of Radiology, Heidelberg,
Baden-Württemberg, Germany
Anahita Fathi Kazerooni
Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia,
PA, United States
Ana R. Fouto
Institute for Systems and Robotics and Department of Bioengineering, Instituto Superior Técnico,
Universidade de Lisboa, Lisbon, Portugal
Andreia S. Gaspar
Felix Glang
Magnetic Resonance Center, Max Planck Institute for Biological Cybernetics, Tübingen, Germany
Matthias Günther
Fraunhofer MEVIS, Bremen, Germany
University Bremen, Bremen, Germany
mediri GmbH, Heidelberg, Germany
Hua Guo
Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of
Medicine, Tsinghua University, Beijing, China
E. Mark Haacke
Department of Radiology, Wayne State University School of Medicine, Detroit, MI, United States
Department of Neurology, Wayne State University School of Medicine, Detroit, MI, United States
Gilbert Hangel
Eva Heckova
Gunther Helms
Department of Clinical Sciences Lund (IKVL), Lund University, Lund, Sweden
Kai Herz
Department of Biomedical Magnetic Resonance, University of Tübingen, Tübingen, Germany
xii List of contributors
Lukas Hingerl
Kevin M. Johnson
Departments of Medical Physics and Radiology, University of Wisconsin – Madison, Madison, WI,
United States
Florian Knoll
NYU School of Medicine, Department of Radiology, New York, NY, United States
FAU Department of Artificial Intelligence in Biomedical Engineering, Erlangen, Germany
Mark E. Ladd
German Cancer Research Center (DKFZ), Division of Medical Physics in Radiology, Heidelberg,
Phil Lee
United States
Department of Radiology, University of Kansas Medical Center, Kansas City, KS, United States
Alexander Leemans
PROVIDI Lab, Image Sciences Institute, University Medical Centre Utrecht, Utrecht,
The Netherlands
Dan Ma
Case Western Reserve University, Cleveland, OH, United States
Karin Markenroth Bloch

Lund University, Lund, Sweden
Zachary Miller
Department of Radiology, University of Washington, Seattle, WA, United States
Mahmud Mossa-Basha
Jeanette A. Mumford
Stanford University, Stanford, CA, United States
Rita G. Nunes
List of contributors xiii
Kristin P. O’Grady
Vanderbilt University Institute of Imaging Science, Vanderbilt University Medical Center,
Nashville, TN, United States
Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center,
Gülin Öz
Center for Magnetic Resonance Research, Department of Radiology, University of Minnesota,
Minneapolis, MN, United States
Daniel Paech
German Cancer Research Center (DKFZ), Division of Radiology, Heidelberg,
Jonathan R. Polimeni
Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital,
Charlestown, MA, United States
Department of Radiology, Harvard Medical School, Boston, MA, United States
Division of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge,
MA, United States
Benedikt A. Poser
Maastricht University, Maastricht, The Netherlands
Sudarshan Ragunathan
Division of Neuroimaging Research, Barrow Neurological Institute, Phoenix, AZ, United States
Jonas Richiardi
Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne,
Switzerland
Stefan Ropele
Department of Neurology, Medical University of Graz, Graz, Austria
Sepideh Sadaghiani
Psychology Department, Beckman Institute for Advanced Science and Technology, University of
Illinois at Urbana-Champaign, Urbana, IL, United States
Seth A. Smith
Vanderbilt University Institute of Imaging Science, Vanderbilt University Medical Center,
Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center,
Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States
Department of Ophthalmology, Vanderbilt University Medical Center, Nashville, TN, United States
xiv List of contributors
Tony Stöcker
German Center for Neurodegenerative Diseases (DZNE), Bonn, Germany
Bernhard Strasser
Carole Helene Sudre
MRC Unit for Lifelong Health and Ageing at UCL, University College London, London,
United Kingdom
Centre for Medical Image Computing, University College London, London, United Kingdom
School of Biomedical Engineering and Imaging Sciences, King’s College London, London,
United Kingdom
Ivan Tkáč
Center for Magnetic Resonance Research, Department of Radiology, University of Minnesota,
Minneapolis, MN, United States
André van der Kouwe
Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital,
Charlestown, MA, United States
Eric K. van Staalduinen
Joost Verhoeff
Department of Radiation Oncology, University Medical Centre Utrecht, Utrecht, The Netherlands
Luca Vizioli
Center for Magnetic Resonance Research (CMRR), Department of Radiology, University of
Minnesota, Minneapolis, MN, United States
Department of Neurosurgery, University of Minnesota, Minneapolis, MN, United States
Natalie L. Voets
Wellcome Centre for Integrative Neuroimaging (FMRIB-WIN), John Radcliffe Hospital, University
of Oxford, Oxford, United Kingdom
Kevin Yuqi Wang
Essa Yacoub
Center for Magnetic Resonance Research (CMRR), Department of Radiology, University of
Minnesota, Minneapolis, MN, United States
Chun Yuan
List of contributors xv
Greg Zaharchuk
Moritz Zaiss
Department Neuroradiology, University Clinic Erlangen, Friedrich Alexander University
Erlangen-Nürnberg, Erlangen, Germany
Hui Zhang
Centre for Medical Image Computing and Department of Computer Science, UCL, London,
United Kingdom
Wei Zhang
Department of Radiology, Washington University School of Medicine, Washington University in
St. Louis, St. Louis, MO, United States
Zechen Zhou
Philips Research North America, Cambridge, MA, United States
Preface
This book arose principally from a highly successful workshop titled “Advanced Neuro MR: Best
Practices for Technical Implementation” that was organized by the International Society for Magnetic
Resonance in Medicine, which took place in Seoul, South Korea, in 2018. It was clear from the work-
shop that there was a desire to summarize state-of-the-art neuroimaging in the form of an edited book.
Many of the speakers at that workshop agreed to contribute to the book, and more experts were recruited
to the effort. The result is a comprehensive summary of the field, provided by its leading researchers.
We are deeply honored and grateful that such distinguished scientists and clinicians have contributed
to the project.
The book is divided into twelve parts, covering a wide range of topics intended to be accessible to
both experts and newcomers. Basic scientists involved in the methodology of neuro MR should find
the book as equally valuable as should technically-oriented clinicians and neuroscientists who seek a
deeper understanding of its principles. After beginning with introductory chapters on fast and robust
imaging, and approaches to image analysis, the book focuses on advanced structural, functional, and
physiological imaging, before covering other specialist MRI contrasts, such as magnetic susceptibil-
ity imaging, magnetization transfer, and relaxometry mapping. The book also covers neurovascular
imaging and advanced magnetic resonance spectroscopy, and then summarizes the opportunities, chal-
lenges, and applications of ultra-high-field magnetic resonance. Please note that there is accompanying
electronic content available for some chapters at: https://www.elsevier.com/books-and-journals/book-
companion/9780128224793.
We would like to express our deepest gratitude to our authors for keeping to a rigorous submission
schedule, despite the huge disruptions to their laboratories and personal lives caused by the global
COVID-19 pandemic. We would also like to thank the Commissioning Editor at Elsevier, Tim Pitts,
the Editorial Project Manager, Emily Thomson, and the Production Project Manager, Prem Kumar
Kaliamoorthi, for their support and guidance from the book’s conception to the final product. We hope
that readers will find this a useful addition to the literature, and a valuable reference tool in their work.
In-Young Choi, Kansas City, KS, United States

Peter Jezzard, Oxford, United Kingdom
xvii
PART
Fast and robust

imaging
1
CHAPTER
Recommendations for neuro MRI

acquisition strategies
1
Tony Stöcker
German Center for Neurodegenerative Diseases (DZNE), Bonn, Germany
1.1 MRI hardware

Since the advent of MRI, the underlying system hardware has been substantially improved. MR image
acquisitions have become faster, and at the same time the image quality has become better. Four factors
were mainly responsible for these improvements: i) an increase of magnetic field strength (3 tesla or
even 7 tesla) provides higher signal-to-noise ratio (SNR), ii) faster switching and more powerful gradi-
ent systems increase the speed of spatial encoding, iii) multi-channel receiver arrays increase SNR and
also enable faster spatial encoding (“Parallel Imaging” – PI), and finally iv) faster and more powerful
computers improve the image reconstruction quality. Based on these hardware improvements, novel
imaging methods were developed which gave access to fast acquisitions of soft-tissue contrasts with
high potential for neuroimaging application. The recommendations given here will focus on MR neu-
roimaging techniques at 3 tesla (3T). This is the current de-facto standard for many recent large-scale
neuroimaging studies. 3T MRI scanners are widely available in the research landscape and, additionally,
provide a good compromise between sensitivity and imaging artifacts. The latter typically also scale up
with magnetic field due to susceptibility-induced inhomogeneity of the main field and inhomogeneity
of the transmit radio frequency (RF) field, which scales with field-dependent wavelength. The latter
effect becomes especially problematic at 7T and will be discussed in Chapter 35 of this book. The
wavelength in the human body is about 30 cm at 3T, and therefore wave-effects in neuro MRI exist, but
do not significantly compromise the image quality. Nevertheless, they need to be taken into account, es-
pecially when deriving quantitative tissue properties. Translating an optimized image acquisition from
one field strength to another is a difficult task. Even if imaging artifacts are neglected, a theoretical or
experimental SNR and CNR (contrast-to-noise ratio) analysis is cumbersome: many variables need to
be considered, such as the sequence timing, tissue relaxation times, and the receiver coil properties. An
excellent example of a detailed SNR & CNR analysis for 3T, 7T, and 9.4T MRI was given by Pohmann
et al. (2016). Nevertheless, a crude estimate can be quickly performed by assuming that the SNR is
proportional to the field strength, the voxel-volume, and the square-root of the acquisition time (due to
incoherent noise cancellation). Adjusting a 3T MRI protocol with isotropic resolution to a 1.5T scanner
with the goal
√ of maintaining the SNR requires either an increase in voxel size (the edge length) by a
factor of 3 2 ≈1.26—while keeping the scan time constant —or an increase in acquisition time by a
factor of 4, while keeping the resolution constant. In that way the recommendations of this chapter can
be roughly extrapolated to other field strengths than 3T.
The other hardware components—the gradient system, the receiver arrays, and the image recon-
struction computer—all have application-specific impact in neuro MR. A powerful gradient system is
Advances in Magnetic Resonance Technology and Applications, Volume 4, ISSN 2666-9099. https://doi.org/10.1016/B978-0-12-822479-3.00009-9
3
4 Chapter 1 Recommendations for neuro MRI acquisition strategies
especially important for diffusion MRI. Typical MRI scanners have a maximum gradient strength of 40
mT/m (“milli-tesla-per-meter”), whereas some advanced products double that performance. If diffusion
MRI is key to answering a research question, then an MRI scanner with high-performance gradients is
recommended. If such a system provides a two-fold SNR gain for a dedicated diffusion MRI protocol,
it saves a factor of 4 in scan time. Therefore higher investments for such a system may dramatically
reduce scan time costs in subsequent studies. Today, phased array head coils with 32 (or more) in-
dividual elements are common. They can provide a substantial acquisition speed-up. The common
technology provides PI acceleration factors between 2–8. PI reduces the acquisition time, and therefore
the SNR. Thus PI acceleration capabilities strongly depend on the image resolution. PI reconstructions
with large acceleration factors on low SNR data are computationally expensive. The performance of
the image reconstruction computer is essential for high-throughput data acquisition. If the computer
is busy reconstructing previously acquired data, execution of the next sequence(s) of the MR protocol
is delayed. High-performance-computing extensions of the hardware may help, but are not considered
for the recommendations in this chapter. Here, the description is based on the assumption that image
acquisition and image reconstruction are performed on the MRI system to guarantee robust and high-
throughput examinations. Subsequently, imaging data is exported to an external database for analysis. It
should be noted that this is not easily possible with MRI raw data, i.e., the complex signals per receiver
channel. Commercially available MRI scanners may have no functionality for realtime raw data export.
Even if this would be the case, the huge size of MRI raw data can generate substantial problems. For
example, a one hour MRI protocol can easily generate 1 Petabyte of raw data. Conversely, the smaller
sized imaging data can be more easily stored and subsequently analyzed.
1.2 From signals to biomarkers

The richness and variety of neuro MR methods is impressive: information on biophysical tissue prop-
erties can be encoded in many ways into the MR signal. The MRI sequence first generates these signals
by weighting the spin magnetization with the property of interest, then encodes spatial information into
the signal, and finally acquires the raw data, which is transferred to the image reconstruction computer.
The reconstructed images have an arbitrary intensity scaling. It is the task of image analysis to derive
quantitative information from the images, which may serve as a potential biomarker of a specific normal
biological or pathogenic process in the brain. Overall, many steps are involved from the generation of
the MR signal to an image-derived biomarker of brain structure and function. These steps are highly in-
terdependent: image acquisition, reconstruction, and analysis capabilities determine each other. Novel
approaches try to combine image reconstruction and analysis into a single process, directly delivering
the biomarker from the MRI raw data. This has consequences for the acquisition.
MRI-derived quantitative tissue properties have limited precision and accuracy. The precision de-
scribes the statistical error in the data, and the accuracy quantifies the acquisition bias. To a large extent
the precision is determined by SNR and CNR, but image artifacts, e.g., induced by involuntary head
motion, also degrade the precision. If the precision is low, image-derived quantities cannot serve as a
robust biomarker. The precision can be estimated in test-retest experiments. If unknown, it is recom-
mended to design a pilot study to estimate the precision. However, even if the precision is known, the
effect size of the biological mechanism of interest is often unknown. A good estimate of the effect
size requires studies with large sample size, including hundreds to thousands of subjects, which is not
1.3 Spatial encoding strategies 5
the case for many studies in human neuroscience, as noted by Button et al. (2013). Therefore a sound
power analysis in advance of a new study is difficult. Often, the required sample size is estimated from
previous knowledge. This is difficult due to the limited accuracy of MRI-derived quantities. In general,
MRI introduces a measurement bias: often it is intended to weight the MR signal to a single tissue
property, but other weightings are unavoidably present. For instance, the spatial encoding process al-
ways encodes—to a small extent—a diffusion weighting. If such effects are not built into the physical
model of the quantity of interest, the fitting is biased. It is computationally too expensive to simulta-
neously account for all physical effects in the acquisition process. However, by changing experimental
conditions, e.g., by adjusting the sequence timing parameters or by examining various MR phantoms
with well-defined properties, it is possible to make predictions about the accuracy of a specific method.
Fortunately, exact knowledge of the accuracy is not always required. If only the percentage change of a
quantity in two experimental conditions is relevant for the interpretation, then meaningful conclusions
may be drawn without knowledge of the accuracy. Still, the change will be biased by the acquisition
process. Reproducibility is therefore only possible if the experimental conditions are maintained. For
this reason, there is a strong need for harmonization of neuro MR protocols. The closer the experimental
conditions between different studies, the better the results can be compared. On the other hand, harmo-
nization of protocols often means “freezing” of protocols potentially over a long time, and therefore
new methodological advances of image acquisition cannot be taken into account. Typically, a trade-off
between efficiency (fast acquisition with high precision) and matureness (well-known accuracy and
good reproducibility) has to be made. This consideration is a general important recommendation for
neuro MRI acquisition strategies, which was taken into account for the protocol proposals presented in
this chapter.
1.3 Spatial encoding strategies

The human brain is a highly structured and complex organ. The highly folded gyri and sulci form the
gray matter of the human cortex. Its regions are connected by short- and long-ranging fiber bundles of
axons, which form the white matter. Imaging the brain with MRI provides spatially resolved informa-
tion on the scale of approximately a millimeter (the voxel size). Acquisition protocols can be adapted to
investigate a specific region of interest (ROI) with higher resolution, but generally whole-brain imaging
is desired: it enables the study of systemic brain structure and function, and it provides the possibil-
ity to compare tissue markers between different brain regions. For example, utilizing the hemispheric
symmetry, tissue composition in a lesion can be compared to the counterpart region in the brain.
Since there is no preferred slice orientation for whole-brain MRI, isotropic spatial image resolution
is generally advisable. For resolutions of one millimeter and below, this is preferably obtained with
3D spatial encoding. Here, the MRI sequence does not perform slice selection, but uses RF pulses that
excite the whole human head (or a large slab through the head). The resulting MR signal is spatially
encoded utilizing gradient waveforms, which define an appropriate trajectory in 3D k-space. The sam-
pling points along the trajectory define the grid of the 3D Fourier transform for image reconstruction.
There are infinite ways to traverse k-space, e.g., curved non-Cartesian trajectories are time-efficient.
However, then the reconstruction problem is complicated. Thus most MRI sequences provided on
commercial MRI systems employ Cartesian k-space encoding. This has several advantages: the image
reconstruction problem is computationally less demanding and, additionally, minor gradient hardware
imperfections typically have little or no effect on the final image. Moreover, a single generic image
reconstruction algorithm can handle the raw data of arbitrary sequences. State-of-the-art Cartesian im-
age reconstruction, including parallel imaging capabilities, are available on all modern MRI scanners.
Therefore the recommendations of this chapter are confined to Cartesian acquisitions. However, the
field of advanced image reconstruction capable by non-Cartesian MRI is rapidly evolving. Recent de-
velopments strongly benefited from deep learning, which improves reconstruction quality and reduces
computation times dramatically. However, those methods are not standard and are more difficult to
apply. An overview of recent developments is given in Chapter 2.
The gradient axes of 3D Cartesian MRI are given by the frequency-encoding axis and two phase
encoding axes. The data are acquired simultaneously with the frequency-encoding gradient (or readout
gradient), providing a fast sampling of k-space in this direction. The sampling along the two “spin-
warp” phase encoding (PE) dimensions is repeated in a two-fold loop structure of the MRI sequence,
which results in long acquisition times. The process can be substantially accelerated with parallel
imaging, which under-samples k-space along the PE dimensions. Under-sampling is a violation of
the Nyquist–Shannon sampling theorem and results in aliasing artifacts, the well-known fold-over arti-
facts in MRI. Parallel imaging (PI) removes (or avoids) these artifacts by utilizing the spatial sensitivity
information of a multi-channel receiver array coil. Modern head coils with 32 or more channels have
sensitivity variation along all three spatial dimensions. Therefore PI acceleration can be performed for
both PE dimensions. 2D-PI is an important feature of 3D spatial encoding, and it can be well exploited
in human head imaging. As mentioned, PI reduces the SNR due to the reduced sampling duration.
Additionally, the noise is amplified by the g-factor (g ≥ 1), which depends on the coil geometry and
the encoding scheme (Pruessmann et al., 1999). Regions with large fold-over have a large g-factor. It
can be substantially reduced using CAIPIRINHA sampling (Breuer et al., 2005), or for short “CAIPI”
sampling. Shifting the under-sampling artifacts relative to one another for the two PE dimensions re-
duces fold-over, and therefore the g-factor. With CAIPI sampling, higher acceleration factors can be
employed without sacrificing image quality.
There exist three more techniques to reduce the spatial encoding time of Cartesian acquisitions: par-
tial Fourier (PF), elliptical sampling (ES), and echo planar imaging (EPI). All methods can be combined
with each other and with PI. Partial Fourier takes advantage of the Hermitian symmetry in k-space,
enabling a reduced sampling in one PE dimension, which can save 10–40% of scan time. Elliptical
sampling (Bernstein et al., 2001) omits sampling of 2D-PE k-space outside an ellipsoid (a circle, in
the case of isotropic spatial encoding), which defines the imaging resolution. Elliptical sampling saves
approximately 25% of scan time and can be applied to many 3D neuro MRI acquisitions. EPI encodes
multiple readouts after a single excitation with alternating frequency-encoding gradients and small PE
gradient blips in between. EPI was introduced as a 2D single-shot method, acquiring a whole-slice after
one RF excitation. The method can be extended to 3D-EPI with two PE dimensions, where for instance
one PE dimension is encoded in the EPI train and the other PE dimension is segmented in an outer
loop. However, segmenting both PE dimensions with respect to the excitation provides more flexibility
for 2D-PI acceleration. Combining a double-segmented 3D-EPI acquisition with CAIPI sampling was
recently introduced as skipped-CAIPI sampling (Stirnberg and Stöcker, 2021). Using the method with
moderate EPI-factors (< 10) is, for instance, well-suited for high-resolution anatomical MRI.
Examples of efficient 2D phase encoding patterns are depicted in Fig. 1.1, where A and B form
the basis for most of the neuro MR applications presented in the remaining sections of this chapter.
1.3 Spatial encoding strategies 7
Fig. 1.1A shows two-fold under-sampling with CAIPI shift (green squares are not acquired) and ellip-
tical sampling. The encoding scheme is well-suited for 3D anatomical T1 -weighted and T2 -weighted
sequences with a two-fold loop structure, where the outer loop encodes contrast and the inner loop
samples a constant number of PE steps. The three red lines in Fig. 1.1A, show inner loop encodings
for the first, middle, and last step of the outer loop. Fig. 1.1B shows the extension to double-segmented
EPI acquisition with skipped-CAIPI sampling. The example depicts a 6-fold PI acceleration of 3D-EPI
with 2D-PE segmentation. The yellow squares connected by yellow arrows are acquired in the first EPI
train, the white squares connected by white arrows are acquired in the second echo train, and the pro-
cess is repeated until all data are acquired. An example beyond Cartesian acquisitions is finally given in
Fig. 1.1C, which shows the wave-CAIPI acquisition (Bilgic et al., 2015). Here, a corkscrew-trajectory
along the readout dimension is utilized in addition to the 2D phase-encoding under-sampling. As op-
posed to Figs. 1.1A and B, this approach takes additional advantage of receiver sensitivity variation in
the readout dimension, which reduces the g-factor and enables higher acceleration factors. However,
3D k-space sampling is no longer on a Cartesian grid, and therefore a dedicated image reconstruction
algorithm is required.
In conclusion, 3D Cartesian acquisitions utilizing 2D parallel imaging with CAIPI sampling are
well suited for rapid whole-brain neuroimaging. The encoding can be combined with other acceleration
techniques, such as partial Fourier imaging, elliptical sampling, and EPI. Such methods are available
on modern MRI systems and enable fast and robust neuro MRI applications.
FIGURE 1.1
Parallel imaging (PI) acceleration schemes to optimally exploit the spatial sensitivity variation of modern MRI re-
ceiver head coils. (A) 2D phase encoding (PE) pattern with two-fold CAIPI under-sampling (green squares are not
acquired) and elliptical sampling. (B) Extension to segmented 3D-EPI acquisitions, which acquire multiple PE steps
per excitation. (C) Further extension to wave-CAIPI acquisitions, which additionally exploit receiver sensitivity
variation in the readout dimension, however, at the expense of departing from the Cartesian grid.
1.4 Large-scale population imaging

A dedicated MR protocol for a single neuro MR method could last from a few minutes up to several
hours, depending on the desired resolution and image contrast. Scanning human subjects limits the total
scan time to approximately one hour; longer examinations may result in subject discomfort and strong
motion artifacts. If a research question is linked to a special technique, the protocol can be designed
accordingly. However, often the combination of complementary information from multiple contrasts
increases the overall sensitivity. Therefore, a versatile MR protocol tries to combine several methods,
each with a short scan time of several minutes, but still providing sufficient resolution and sensitivity. It
is a good choice to use a well-tested MR-protocol from someone else, if it meets your own requirements.
There exist several large-scale population imaging studies that developed multi-purpose MRI protocols.
Using such a protocol has two advantages. Firstly, these protocols were thoroughly designed and tested.
Secondly, newly acquired imaging data is compatible with the population study and results can be more
easily correlated.
Different approaches exist to gather large-scale neuro MRI data. Retrospective imaging studies per-
form meta-analyses on large amounts of neuroimaging data across many different studies and sites; a
prominent example is the ENIGMA study (http://enigma.ini.usc.edu/). Conversely, a prospective study
collects data with identical set-up and methods tailored to the study questions. Usually, the MRI pro-
tocols of these studies have been extensively piloted. Table 1.1 gives an overview of four prospective
population imaging studies. The original human connectome project (HCP) was the first large-scale
imaging study to apply advanced MRI technology in more than 1000 healthy young adults. High per-
formance gradient systems, as presently available in commercial MRI scanners, were developed in the
context of the HCP. Meanwhile, several HCP sub-studies were launched, each addressing specific neu-
roscientific research questions, and each with a dedicated MRI protocol. Another approach is given
by multi-center prospective studies, where imaging protocols are harmonized across many different
sites and scanner types, aiming at increased sample size. A successful example of this approach is the
Alzheimer’s disease neuroimaging initiative (ADNI). The initial five-year study (ADNI-1) was twice
extended (ADNI-2, and ADNI-3), where each time the MR protocols were adjusted to the newest
technology, given by the common denominator available at the participating sites. There have been
increasing efforts to utilize advanced MRI technology in large-scale population studies by means of
dedicated centers, equipped with MRI systems exclusively for the study. The UK-Biobank study utilizes
a 30 minutes neuro MRI protocol, with a target population of 100,000 subjects. A similar concept, with
the additional goal of longitudinal examination, is implemented by the Rhineland study. The study uti-
lizes a 60 minutes neuro MRI protocol obtained on two MRI systems equipped with high-performance
gradients (80 mT/m). The study aims to include 20,000 subjects with re-examination every 3–4 years.
All studies of Table 1.1 utilize 3T MRI systems. Detailed information on the MRI protocols is available
on the respective web-sites. The next section outlines the one-hour MR protocol of the Rhineland study
in more detail.
1.5 Example multi-purpose protocols

This section presents two example MR protocol proposals for multi-purpose neuroimaging applica-
tions. Imaging examples will be given in the next section, including more details on the individual
1.5 Example multi-purpose protocols 9
Table 1.1 Large-scale prospective population imaging studies with dedicated neuro MRI proto-
cols. Detailed study information, including the MRI protocols can be obtained from the study
web-sites.
Study Name General Information
- original HCP (2009–2014): MRI with custom gradient hardware
Human Connectome Project (HCP) - 4 substudies (since 2015) on aging and developing brain
- sample size / substudy: N ≈ 500–1,500
- https://humanconnectome.org
- multi-center study in patient population (MCI & AD)
Alzheimers Disease Neuroimaging - active since 2009, three study extensions (current: ADNI-3)
Initiative (ADNI) - sample size: N ≈ 2,000
- http://adni.loni.ucla.edu
- 4 dedicated study centers / MR-protocol inspired by HCP
- brain, cardiac, & abdominal MRI / 30 min brain protocol
UK Biobank Imaging Study - current/target N ≈ 25,000/100,000 (re-invite ≈ 10,000)
- subjects recruited from the UK Biobank cohort (≈ 500,000)
- https://www.ukbiobank.ac.uk
- 2 dedicated study centers / MRI with high-performance gradients
- 8 h phenotyping per subject, including 60 min brain MRI protocol
Rhineland Study - current/target N ≈ 5,000/20,000
- longitudinal study (rescan every 3–4 years)
- https://www.rheinland-studie.de (in German)
sequences and methods. Table 1.2A shows an advanced research protocol, which is (except for a miss-
ing body-fat exam) exactly the MRI-protocol of the Rhineland study. The total scan time is below one
hour per subject. All sequences, except #1 and #8, employ 3D-acquisition with 2D acceleration and, if
possible, elliptical sampling. Sequence #8, DWI, requires a different acceleration strategy, as will be
explained in the next section. The first sequence is a short 3D scout, enabling automatic slice position-
ing of all subsequent acquisitions. The second sequence, 3DREAM (Ehses et al., 2019), acquires fast
whole-brain B0 and B1 field maps, providing off-resonance and actual flip angle distribution. These
maps are required for various calibration purposes in the analysis, e.g., for quantitative tissue parame-
ter mapping. With the current available technology, it is not feasible to acquire all possible neuro MR
contrasts of interest in one hour and, at the same time, high image quality and resolution. Therefore, the
protocol is designed in two parts: the first 8 sequences form the core protocol with 45 minutes net scan
time. Contrasts for the core protocol are assumed to be of higher relevance and will be acquired for every
subject of the study. The second part is the free protocol. It consists of promising applications, however
only one of them is acquired per subject. This keeps the total examination time below 1 h, and results
in a smaller sub-population for each contrast of the free protocol. The specific choices for the core and
free protocol were based on the general objectives of the Rhineland study. Since it is a long-term study,
it might be possible that objectives will adapt, and therefore also the protocol will adapt. A sequence
may move from the core protocol to the free protocol, or vice versa, or could be completely removed.
Likewise, new emerging techniques may enter the free protocol. (The free protocol of the Rhineland
study already uses more specific sequences, e.g., for hippocampal subfield imaging or T1ρ mapping.)
This acquisition strategy was considered to be both forward-looking and at the same time mature. It
may serve as a recommendation for upcoming long-term studies. Finally, Table 1.2B presents a fast
research protocol proposal, providing high-quality neuroimaging data in less than 10 minutes, which
is well-suited for a clinical setting. The proposal includes fast perfusion imaging, which might be of
interest in a patient population. Note that the anatomical images are acquired with 1.0 mm isotropic
resolution, which is still the common standard in many advanced neuroimaging studies. The short scan
times result from higher acceleration factors, which can be achieved with 2D acceleration.
Table 1.2 Example recommendations for neuroimaging protocols. All sequences provide whole-
brain coverage with isotropic resolution. A) The research protocol of the Rhineland study, pro-
viding many promising MRI contrasts in a short scan time. The bottom rows printed in gray font
form the optional free protocol (FP). B) Fast research protocol to acquire high-quality neuroimag-
ing data in less than 10 minutes for several clinically relevant contrasts.
A
# Contrast Sequence Name Resolution [mm3 ] PI Accel. TA [min] Main Features
& Matrix Size
1 Scout GRE 1.6 × 1.6 × 1.6 3×1 00:14 auto align
160 × 160 × 128
2 field mapping 3DREAM 5.0 × 5.0 × 5.0 2×21 00:10 B1 & B0 maps
54 × 54 × 48
3 rs-fMRI 3D-EPI 2.4 × 2.4 × 2.4 1×62 10:25 TE/TR=30/570 ms
90 × 90 × 60 1070 volumes
4 T1 -weighted ME- 0.8 × 0.8 × 0.8 1×21 06:35 4×TE: 1.6–6.5 ms
MPRAGE 320 × 320 × 224 elliptical sampling
5 T2 -weighted 3D-TSE 0.8 × 0.8 × 0.8 1×21 04:47 external ACS, PF=6/8
320 × 320 × 224 elliptical sampling
6 FLAIR IR-3D-TSE 1.0 × 1.0 × 1.0 1×21 04:37 ext. ACS, ellipt. sampl.
256 × 256 × 176
7 QSM ME-SC- 0.8 × 0.8 × 0.8 3×21 06:12 maps: ω, χ , T2∗
3D-EPI 270 × 270 × 176 6×TE: 7–32 ms
EF=7, avg=4, PF=6/8
8 DWI (CS-DSI) SMS-DW- .5 × 1.5 × 1.5 MB 3 12:04 117 q-space samples
SE-EPI 140 × 140 × 93 bmax =6800 s/mm2
TE=107 ms, PF=6/8
45:04
FP1 MPM SC-3D-EPI 0.8 × 0.8 × 0.8 3×21 7:15 maps: PD, T1 , MTsat
270 × 270 × 176 EF=7, PF=6/8, avg=5
FP2 Perfusion te-PCASL 3.3 × 3.3 × 3.3 1×21 7:13 8×8 Hadamard time-enc.
3D-GRASE 64 × 64 × 36 CBF+ATT maps
EF=32, avg=7
FP3 CEST ss-CEST- 2.0 × 2.0 × 2.0 1×62 6:56 45 freq. offsets, 2 B1 amp.
3D-EPI 108 × 108 × 70 EF=21, PF=6/8
··· ···
continued on next page
1.6 Acquisition of neuro MRI contrasts 11
Table 1.2 (continued)

B
# Contrast Sequence Name Resolution [mm3 ] PI Accel. TA [min] Main Features
& Matrix Size
1 Scout GRE 1.6 × 1.6 × 1.6 3×1 00:14 auto align
4 T1 -weighted ME-MPRAGE 1.0 × 1.0 × 1.0 1×31 02:40 4×TE: 1.6-6.5 ms
5 T2 -weighted 3D-TSE 1.0 × 1.0 × 1.0 2×21 01:47 external ACS, PF=6/8
8 DWI (DTI) SMS-DW-SE-EPI 2.0 × 2.0 × 2.0 MB 3 2:14 33 directions
108 × 108 × 69 TE=76 ms, b=1000 s/mm2
7 Perfusion te-PCASL 3.3 × 3.3 × 3.3 1×21 02:12 CBF map (single PLD)
3D-GRASE 64 × 64 × 36 EF=32, avg=6
09:05
1.6 Acquisition of neuro MRI contrasts

Multi-purpose neuro MRI exams acquire several contrasts, providing complementary information on
brain structure and function. Some MR techniques even provide quantitative information on the bio-
physical tissue composition inside the voxel. This section briefly highlights examples of the most
important techniques. All methods will be discussed in greater depth in other chapters of this book.
The images shown in this section were obtained with the MR protocol presented in Table 1.2A.
1.6.1 Brain anatomy

Imaging brain anatomy with high resolution is a key component for many neuro MRI exams. It serves as
the basis for segmenting brain structures and performing morphometric analysis, e.g., to quantify brain
atrophy in neurodegeneration. T1 -weighted 3D sequences, such as the MP-RAGE sequence, provide
excellent image quality and contrast for brain segmentation. The sequence utilizes a global inversion in
an outer loop, followed by an inner loop with a gradient-echo readout. Originally, the MP-RAGE loop
structure was fixed to the phase encoding dimensions. However, the approach can be easily combined
with 2D-CAIPI acceleration and elliptical sampling, which provides significant speed-up without sac-
rificing image quality (Brenner et al., 2014). High-resolution whole-brain imaging of brain anatomy
with T2 -weighting can be efficiently obtained with a 3D turbo-spin echo sequence (3D-TSE). Again,
2D acceleration with elliptical sampling can be efficiently applied to the 2-fold loop structure of 3D-
TSE (Busse et al., 2008). However, the strong signal from the eyes in T2 -weighted scans may lead to PI
artifacts if the PI auto-calibration signal (ACS) is acquired during phases of eye motion. Therefore it is
advisable to acquire the ACS in a fast prescan. The T2 -weighted images are less suited for brain segmen-
tation, but they provide contrast for tissue boundaries which are difficult to detect with T1 -weighting,
e.g., the border from gray matter to the dura. A combined segmentation requires both scans at the same
resolution. With the aforementioned acquisition methods, both scans can be acquired with 0.8 mm
resolution and good image quality in about 10 minutes. Fig. 1.2 depicts anatomical images acquired
with the MR protocol of the Rhineland study. Here, one important detail is the bandwidth-matching
between MP-RAGE and 3D-TSE. Typically, MP-RAGE has a lower readout encoding bandwidth than
3D-TSE, which results in subtle image artifacts and geometric distortions close to areas of tissue-air
interfaces. Acquiring multiple gradient-echoes at higher bandwidth (ME-MP-RAGE) with subsequent
signal-averaging maintains acquisition time and CNR and, additionally, provides a perfect registration
to 3D-TSE (van der Kouwe et al., 2008).
FLAIR images are often acquired in clinical neuro MRI. FLAIR provides T2 -weighting with CSF
signal nulling, obtained by applying a global inversion pulse before each excitation pulse of the 3D-
TSE sequence. The contrast is well-suited for detecting white matter lesions (WM hyper-intensities).
Although presumably possible, so far it has not been shown that the FLAIR contrast can be artificially
generated from T1 - and T2 -weighted acquisitions. Therefore, FLAIR is part of the Rhineland study
core protocol. It might be removed in the future if, for instance, artificial intelligence (AI) applications
enable robust and fail-safe FLAIR predictions from the other anatomical scans. A FLAIR example is
depicted in Fig. 1.2C. The image has 1.0 mm isotropic resolution with a total scan time of 4:37 min.
Due to its reduced SNR, FLAIR cannot be acquired with comparable image quality, resolution, and
scan time as the T1 - and T2 -weighted scans. All three acquisitions of Fig. 1.2 utilize the acceleration
pattern depicted in Fig. 1.1A.
FIGURE 1.2
Examples of 3D whole-brain anatomical scans acquired at 3 tesla, which are well-suited for brain segmentation,
morphometry, and lesion detection. A) T1 -weighting, B) T2 -weighting. C) FLAIR. The corresponding acquisition
protocols are given in Table 1.2A (# 4,5,6).
1.6.2 Tissue microstructure

Modeling tissue microstructure with diffusion-weighted imaging (DWI) requires the acquisition of
many images. Tens to hundreds of repetitions are common, varying the directions and the strength (b-
value) of the diffusion weighting gradients. Chapters 8–11 of this book are dedicated to the acquisition
and reconstruction of DWI data. For most applications the state-of-the-art sequence is SMS-DW-SE-
EPI: a diffusion-weighted spin-echo EPI sequence with simultaneous-multi-slice acquisition. SMS
utilizes multi-band (MB) pulses, which select multiple slices at the same time. As in conventional
PI, the overlapping SMS signals can be separated with knowledge of the coil array receive sensitivi-
ties. Chapter 3 of this book provides a comprehensive introduction to the technique. SMS-EPI provides
a “snapshot” acquisition of the DW-encoding per slice(s), avoiding motion-induced phase errors re-
sulting from the DW-gradients, as they would occur with segmented 3D-EPI. Therefore SMS is the
recommended acceleration technique for DWI. As mentioned in Section 1.1, DWI acquisitions ben-
efit from high-performance gradients. This reduces the duration of the DW-gradients, and therefore
the echo time, TE. Shorter TE leads to a higher SNR, which can be invested in higher resolution. For
standard gradient systems with a maximum amplitude of 40 mT/m, an isotropic resolution of ≈2 mm
is recommended. A gradient amplitude of 80 mT/m enables acquisitions with 1.5 mm isotropic resolu-
tion, as shown in the following examples. The repeated DWI acquisitions sample points in “diffusion
space”, also known as q-space. Each DW direction and b-value defines a point in q-space. The acqui-
sition scheme is often linked to the analysis methods. Robust diffusion tensor imaging (DTI) requires
a single b ≈ 1000 s/mm2 and approximately 30 directions, i.e., sampling a sphere (shell) in q-space,
plus a b = 0 acquisition. Many biophysical microstructure models require acquisitions with varying b-
value. A very successful approach is HARDI acquisition, which samples multiple shells (high angular
diffusion imaging). An equidistant q-space sampling is called diffusion spectrum imaging (DSI). Many
diffusion models can be derived from DSI data, however, the acquisition time is long. Fortunately,
DSI acquisitions can be substantially accelerated by applying compressed sensing (CS) in q-space:
the smooth variation of the q-space signal distribution enables a sparse signal representation. Using
CS, the total distribution can be reconstructed from under-sampled q-space data. Fig. 1.3A shows the
CS-DSI acquisition as utilized in the Rhineland study. 117 randomly distributed q-space samples are
acquired, and CS reconstruction provides the full distribution sampled with 257 points. The corre-
sponding acquisition parameters are given in Table 1.2A (#8), and a detailed discussion of the protocol
was given by Tobisch et al. (2018). Fig. 1.3B depicts some derived measures of tissue microstructure.
The concepts will be discussed in detail in Chapters 8–11 of this book. The top row shows maps from
mathematically inspired diffusion models: fractional anisotropy (FA) and mean diffusivity (MD), from
the tensor model, and the radial kurtosis (RK). The bottom row shows maps from biologically inspired
diffusion models: neurite density (ND) and orientation dispersion (OD) from NODDI, and intra-axonal
restricted volume fraction (FR) from CHARMED. Reliable FR estimates require sampling of high
b-values up to b ≈ 6–7 × 103 s/mm2 . Fig. 1.3C shows results from deterministic fiber tracking, visual-
izing macroscopic structural connectivity of the brain. The tracking is based on voxel-wise estimates of
the orientation distribution function (ODF), which is able to model multiple crossing fibers inside the
voxel (c.f. lower left of Fig. 1.3C).
FIGURE 1.3
Diffusion spectrum imaging (DSI) with a compressed-sensing approach in q-space (A) enables fast acquisition of
DWI data, which is suitable to reconstruct multiple tissue microstructure maps (B), as well as macroscopic brain
structural connectivity or fiber tracking (C). The sequence details are given in Table 1.2A (#8). Figure adapted from
(Tobisch et al., 2018).
1.6.3 The brain at work and rest

Similar to diffusion imaging, functional MRI (fMRI) also acquires a large number of images. The
readout method of choice is gradient-echo EPI. Repeated imaging with T2∗ -weighting samples brain
activity (the BOLD response) at high frame rate. The same sequence can be used to study brain activity
resulting from certain stimuli (task-fMRI) or to investigate the functional connections between different
brain regions in absence of external input (resting-state fMRI). Chapters 15 and 16 of this book provide
a detailed introduction to fMRI data acquisition and analysis. Here, we name four important factors
which determine the BOLD-sensitivity of the EPI acquisition: the echo time (TE), the repetition time
per volume (TR), the total acquisition time (TA), and the temporal SNR (tSNR) of the EPI time series.
At 3 tesla, TE ≈ 30 ms provides maximum BOLD-sensitivity. The tSNR is given by the mean voxel
signal divided by its standard deviation. The tSNR depends on the image SNR, and therefore on the
resolution. An isotropic resolution of 2–3 mm is a common and good choice for fMRI at 3 tesla. More
samples of the time series yield a higher statistical power in the analysis. For instance, using a fixed TA
of 10 min and TR of 600 ms yield 1000 whole-brain images. Such a high sampling rate is also bene-
ficial for identification and removal of physiological noise (nuisance regression), i.e., cleaning signal
variations due to heart beat and breathing, which contaminate the BOLD signal. Therefore, accelerated
EPI acquisitions offer the best advantage. SMS-EPI acquisitions with multi-band acceleration factors
MB ≈ 6–8 are common. This approach is utilized by large population studies, such as HCP and UK-
Biobank. An alternative approach is 3D-EPI with 2D-CAIPI acceleration. Compared to slice-selective
approaches, 3D acquisitions can achieve fat suppression in a faster and more elegant way. Since water-
selective SMS pulses are not efficient, SMS-EPI utilizes standard fat suppression pre-pulses, which are
time-consuming and also slightly influence the water signal. Instead, 3D-EPI can take advantage of
short water-selective pulses, which excite the whole brain or a large slab. Moreover, since 3D-EPI at
short TR operates in a steady state, the flip-angles are reduced (Ernst angle). This reduces the SAR
(specific absorption rate) and makes the sequence particularly well suited for high field applications.
Fig. 1.4 summarizes the results of a comparative study for fMRI acquisitions with SMS and 3D-EPI at
3 tesla (Stirnberg et al., 2017), which was performed in the piloting phase of the Rhineland study. The
sequence diagrams in Fig. 1.4A schematically show the increased TR of SMS-EPI due to the fat sup-
pression (FS) and multi-band (MB) excitation at the beginning, which can be achieved in shorter time
with a water-selective excitation (WE) pulse for 3D-EPI (a binomial 1-1 hard pulse in the example).
Thus, a TR-matched 3D-EPI sequence can utilize lower acceleration factors than SMS. This results in
higher temporal SNR (tSNR) as depicted in Fig. 1.4B, where TR=530 ms and the acceleration factors
were 8 and 6 for SMS and 3D-EPI, respectively. The tSNR analysis in gray matter (Fig. 1.4B, bottom)
shows that the advantage further increases if additional nuisance regression and bandpass filtering is
performed. Consequently, functional networks are more pronounced with 3D-EPI than with SMS-EPI,
as shown in the resting-state analysis depicted in Fig. 1.4C. However, SMS-EPI also provides highly
significant functional networks. In conclusion, both options are well-suited state-of-the-art options for
fMRI at 3 tesla.
FIGURE 1.4
Comparison of functional MRI acquisitions with SMS-EPI and 3D-EPI: A) Sequence diagrams, B) temporal SNR
analysis, C) resting-state network analysis. Details of the 3D-EPI acquisition protocol are given in Table 1.2A (#3).
Figure adapted from (Stirnberg et al., 2017).
1.6.4 Brain perfusion

Constant blood supply provides the oxygen for healthy brain function. The process of blood penetrat-
ing brain tissue, first through larger arteries, and then to a dense network of small capillaries, is called
perfusion. MRI can quantify the process by means of arterial spin labeling (ASL), which labels (typi-
cally inverts) the magnetization of inflowing blood. The T1 -recovery of the MR signal can be measured
and, utilizing kinetic models of brain perfusion, the cerebral blood flow (CBF) can be estimated. ASL
uses blood flow as an endogenous perfusion contrast. An overview of available techniques is given in
Chapter 12 of this book. As always in MRI, a method-specific acquisition bias is present. Since CBF
is a clinically relevant marker, protocol harmonization is especially important to quantitatively com-
pare results. According to a consensus paper (Alsop et al., 2015), the proposed acquisition method is
PCASL (pseudo-continuous ASL) with a 3D-GRASE readout. PCASL performs blood labeling in a
single slice positioned below the brain and covering the major arteries. Labeling durations of a few
seconds provide a stable perfusion weighting. After a post-labeling delay (PLD), whole-brain images
are acquired with a 3D-GRASE readout, i.e., a multiple-spin-echo sequence employing an EPI train
for each echo. The experiment is performed at least twice, with and without labeling, and the differ-
ence image provides the perfusion-weighted signal. Due to the low CNR of perfusion-weighted images,
isotropic voxel-sizes of 3–4 mm are common and signal averaging of multiple repetitions is needed.
If the data are acquired for a single PLD, robust CBF estimates can be drawn from short acquisitions
of 2–3 minutes. In this case, the kinetic model is based on several assumptions about the physiology
and MR tissue parameters, which might not hold over wide age-ranges and different patient groups.
Acquisitions of multiple PLD information provide a more robust and less biased fit of CBF. In addition,
applying a kinetic model to multi-PLD data enables an estimate of arterial transit times (ATT), i.e., the
blood travel time required to reach a certain brain region. The recommended method of choice is so-
called time-encoded PCASL (Teeuwisse et al., 2014), where the PLD is kept constant, but the labeling
duration is varied with different block lengths for the label and control condition. The encoding scheme
follows a Hadamard matrix, which can be inverted to reconstruct high-quality perfusion-weighted data
for multiple PLDs. Moderate PI acceleration is additionally advisable as it reduces motion-induced arti-
facts and allows for the acquisition of more encoding steps for a fixed acquisition time. State-of-the-art
perfusion imaging with a PI-accelerated time-encoded PCASL-3D-GRASE sequence provides stable
estimates of CBF and ATT at approximately 3 mm isotropic resolution in a scan time below 10 minutes.
An important extension of the approach enables probing of the permeability of the blood-brain-barrier
(BBB). Here, the experiment has to be repeated several times with varying T2 -weighting, which can
be conveniently applied during the PLD dead-time (Schmid et al., 2015). Fitting the exponential decay
provides T2 maps of the perfusion-weighted signal. Blood has considerably longer T2 than gray matter.
Therefore the observed decrease of T2 at longer PLDs may be interpreted as water proton transition
from the arterial compartment into the gray matter. A malfunction of the BBB due to disease would
influence the result. The application has huge clinical potential; however, examinations times are still
long. Fig. 1.5 shows examples of perfusion MRI results with a time-encoded PCASL-3D-GRASE se-
quence, as applied in the free protocol of the Rhineland study. Figs. 1.5A and B depict maps of cerebral
blood flow (CBF) and the arterial transit time (ATT), respectively. Fig. 1.5C shows T2 -mapping results
from an acquisition repeating the entire sequence seven times, each with a different T2 -weighting. The
decrease of T2 at late PLDs reflects transition of the labeled water molecules through the BBB. The
total acquisition time for Fig. 1.5C was 45 minutes; however, robust T2 maps may be already obtained
with fewer T2 -weightings (≈4–5).
1.6.5 Biophysical tissue properties

In quantitative MRI (qMRI) a series of weighted images is acquired. Fitting the data to a physical
model enables mapping of biophysical tissue properties. Chapters 22–29 of this book are dedicated
FIGURE 1.5
Whole-brain imaging of brain perfusion with a time-encoded PCASL-3D-GRASE sequence, which provides maps
of cerebral blood flow (CBF) (A), and arterial transit time (ATT) (B). Fig. C depicts blood-brain permeability via
T2 -mapping, obtained from an additional acquisition of multiple T2 -weightings. The acquisition protocol (without
T2 -weighting) is given in Table 1.2A (# FP2).
to the most relevant methods currently applied in neuroimaging. This section presents examples to
rapidly acquire quantitative tissue parameter maps. Quantitative-susceptibility-mapping (QSM) uses
phase images of gradient-echo (GRE) acquisitions to reconstruct the magnetic susceptibility in tissue.
This is, for example, an important marker of increased brain iron load as observed in many neurodegen-
erative diseases. If the GRE sequence acquires multiple echoes with appropriate timing, it is possible
to additionally reconstruct maps of the effective transverse relaxation time, T2∗ , from the magnitude
images. Applying the GRE sequence with appropriate flip angles for T1 - and PD-weighting (proton
density), enables the reconstruction of T1 and PD maps. An additional acquisition with an off-resonant
preparation pulse for MT weighting enables the reconstruction of magnetization transfer saturation,
MTsat , which is a measure of proton magnetization exchange between larger molecules and water.
The acquisition of all contrasts with subsequent reconstruction of the quantitative maps was termed
multi-parameter mapping (MPM) by Weiskopf et al. (2013). The MT acquisition principle can be fur-
ther extended to acquire more information about the molecular environment. The technique is known
as CEST-imaging (chemical exchange saturation transfer). Snapshot-CEST repeats the off-resonant
preparation with a train of preparation pulses, followed by a fast whole-brain GRE readout (Zaiss et
al., 2018). If the acquisition is repeated with varying off-resonances, the data samples the so-called
z-spectrum, a frequency-resolved quantification of proton exchange. The technique is less specific, but
more sensitive than MR spectroscopy, enabling fast imaging of molecular information with isotropic
resolution of ≈2 mm.
All acquisitions can be efficiently obtained with the segmented skipped-CAIPI 3D-EPI (SC-3D-
EPI) readout outlined in Section 1.3. Using substantial double segmentation as well as PI acceleration
along both PE axes reduces the EPI factor, and therefore EPI-specific geometric distortions and other
artifacts induced by field inhomogeneity. In this way, high quality images can be obtained in a short
scan time. Fig. 1.6 shows examples of QSM, MPM, and CEST imaging, acquired with the MR protocol
of the Rhineland study. The snapshot-CEST (ss-CEST) acquisition was repeated twice with varying B1
amplitude to quantify the effect of B1 variation on the z-spectrum. In this context, it has to be noted
that all qMRI reconstructions require information on field inhomogeneities (B0 and B1 maps), which
were obtained with the 3DREAM sequence (Table 1.2A, #2).
FIGURE 1.6
Quantitative MRI reconstructed from whole-brain skipped-CAIPI 3D-EPI acquisitions. A) off-resonance map
(top) and derived quantitative susceptibility map (QSM, bottom). B) Multi-parametric-mapping (MPM) provid-
ing quantitative maps for T1 , T2∗ , PD, and MTsat . C) Contrasts derived from whole z-spectrum CEST imaging:
semisolid magnetization transfer (ssMT, top), and amide proton transfer (APT, bottom). Sequence details are given
in Table 1.2A (#7, FP1, FP3, respectively).
1.7 Conclusions and future prospects

This chapter provided recommendations and examples for neuro MRI acquisition strategies. The dis-
cussion was confined to 3 tesla MRI, which is today the most commonly used system for neuroimaging
research and advanced clinical applications. Many neuro MR applications require whole-brain imag-
ing. Therefore the MRI acquisition should take advantage of 3D spatial encoding with 2D parallel
imaging acceleration, if possible. Cartesian acquisitions are currently still the method of choice, as they
References 19
can be robustly applied in high-throughput studies on commercial MRI systems. Using state-of-the-art
receiver array coils, acceleration factors of 2–8 are possible, depending on the contrast and resolution
of interest. For example structural T1 - and T2 -weighted anatomical imaging with high quality and 0.8
mm isotropic resolution can be obtained in approximately 5–6 minutes, respectively. The recommen-
dations are based on MRI acquisition technologies which are both state-of the-art and widely available
on modern MRI systems. The specific sequences are either available as products by the vendor or can
be obtained via research agreements. The protocol design also has to consider the available image anal-
ysis tools. Ideally, the imaging data should be compatible with existing tools and well-suited for future
analysis concepts.
Given the speed of MRI method development, the advanced MR protocols of today will be outdated
in the (near) future. Recent advances in acquisition technology and image reconstruction will enable
routine use of non-Cartesian acquisitions in the future. Writing the same chapter again in a few years
would potentially name 7T as the current method of choice for neuro MR, enabling even faster ac-
quisitions and better image contrast. Already today, the technology provides many promising contrasts
which can be acquired in a single one-hour multi-purpose exam. As an example, this chapter presented
the MR protocol of the Rhineland study. The methods provide biophysical information of the tissue
composition within a single voxel. Quantitative tissue parameters, microstructure, and molecular in-
formation can be encoded with neuro MRI, all enriching our understanding of the structure and the
function of the human brain. However, the clinical and diagnostic value still needs to be proven for
many of these techniques. This will require more large-scale imaging studies in the general population
and on patients in the future.
References
Alsop, D.C., Detre, J.a., Golay, X., Günther, M., Hendrikse, J., Hernandez-Garcia, L., Lu, H., Macintosh,
B.J., Parkes, L.M., Smits, M., van Osch, M.J.P., Wang, D.J.J., Wong, E.C., Zaharchuk, G., 2015. Recom-
mended implementation of arterial spin-labeled perfusion MRI for clinical applications: a consensus of the
ISMRM perfusion study group and the European consortium for ASL in dementia. Magnetic Resonance in
Medicine (ISSN 1522-2594) 73, 102–116. https://doi.org/10.1002/mrm.25197. http://www.ncbi.nlm.nih.gov/
pubmed/24715426.
Bernstein, M.A., Fain, S.B., Riederer, S.J., 2001. Effect of windowing and zero-filled reconstruction of MRI data
on spatial resolution and acquisition strategy. Journal of Magnetic Resonance Imaging (ISSN 1053-1807) 14
(3), 270–280. https://doi.org/10.1002/jmri.1183.
Bilgic, B., Gagoski, B.a., Cauley, S.F., Fan, A.P., Polimeni, J.R., Grant, P.E., Wald, L.L., Setsompop, K., 2015.
Wave-CAIPI for highly accelerated 3D imaging. Magnetic Resonance in Medicine (ISSN 1522-2594) 73 (6),
2152–2162. https://doi.org/10.1002/mrm.25347. http://www.ncbi.nlm.nih.gov/pubmed/24986223.
Brenner, D., Stirnberg, R., Pracht, E.D., Stöcker, T., 2014. Two-dimensional accelerated MP-RAGE imaging with
flexible linear reordering. Magma (New York, N. Y.) (ISSN 1352-8661) 27 (5), 455–462. https://doi.org/10.
1007/s10334-014-0430-y. http://www.ncbi.nlm.nih.gov/pubmed/24510154.
Breuer, F., Blaimer, M., Heidemann, R.M., Mueller, M.F., Griswold, M., Jakob, P.M., 2005. Controlled aliasing
in parallel imaging results in higher acceleration (CAIPIRINHA) for multi-slice imaging. Magnetic Resonance
in Medicine (ISSN 0740-3194) 53 (3), 684–691. https://doi.org/10.1002/mrm.20401. http://www.ncbi.nlm.nih.
gov/pubmed/15723404.
Busse, R.F., Brau, A.C., Vu, A., Michelich, C.R., Bayram, E., Kijowski, R., Reeder, S.B., Rowley, H.A., 2008.
Effects of refocusing flip angle modulation and view ordering in 3D fast spin echo. Magnetic Resonance in
Medicine (ISSN 1522-2594) 60 (3), 640–649. https://doi.org/10.1002/mrm.21680.
Button, K.S., Ioannidis, J.P.a., Mokrysz, C., Nosek, B.a., Flint, J., Robinson, E.S.J., Munafò, M.R., 2013. Power
failure: why small sample size undermines the reliability of neuroscience. Nature Reviews. Neuroscience
(ISSN 1471-0048) 14 (5), 365–476. https://doi.org/10.1038/nrn3475. http://www.ncbi.nlm.nih.gov/pubmed/
23571845.
Ehses, P., Brenner, D., Stirnberg, R., Pracht, E.D., Stöcker, T., 2019. Whole-brain B1-mapping using three-
dimensional DREAM. Magnetic Resonance in Medicine (ISSN 1522-2594) 82 (3), 924–934. https://doi.org/
10.1002/mrm.27773.
Pohmann, R., Speck, O., Scheffler, K., 2016. Signal-to-noise ratio and MR tissue parameters in human brain
imaging at 3, 7, and 9.4 tesla using current receive coil arrays. Magnetic Resonance in Medicine (ISSN 1522-
2594) 75 (2), 801–809. https://doi.org/10.1002/mrm.25677.
Pruessmann, K.P., Weiger, M., Scheidegger, M.B., Boesiger, P., 1999. SENSE: sensitivity encoding for fast MRI.
Magnetic Resonance in Medicine (ISSN 0740-3194) 42 (5), 952–962. http://www.ncbi.nlm.nih.gov/pubmed/
10542355.
Schmid, S., Teeuwisse, W.M., Lu, H., van Osch, M.J.P., 2015. Time-efficient determination of spin compartments
by time-encoded pCASL T2-relaxation-under-spin-tagging and its application in hemodynamic characterization
of the cerebral border zones. NeuroImage (ISSN 1095-9572) 123, 72–79. https://doi.org/10.1016/j.neuroimage.
2015.08.025.
Stirnberg, R., Stöcker, T., 2021. Segmented K-space blipped-controlled aliasing in parallel imaging for high spa-
tiotemporal resolution EPI. Magnetic Resonance in Medicine 85 (3), 1540–1551. https://doi.org/10.1002/mrm.
28486. https://onlinelibrary.wiley.com/doi/abs/10.1002/mrm.28486.
Stirnberg, R., Huijbers, W., Brenner, D., Poser, B.A., Breteler, M., Stöcker, T., 2017. Rapid whole-brain resting-
state fMRI at 3 Tesla: efficiency-optimized three-dimensional EPI versus repetition time-matched simultaneous-
multi-slice EPI. NeuroImage (ISSN 1053-8119) 163, 81–92. https://doi.org/10.1016/j.neuroimage.2017.08.031.
http://linkinghub.elsevier.com/retrieve/pii/S105381191730678X.
Teeuwisse, W.M., Schmid, S., Ghariq, E., Veer, I.M., Van Osch, M.J., 2014. Time-encoded pseudocontinuous
arterial spin labeling: basic properties and timing strategies for human applications. Magnetic Resonance in
Medicine (ISSN 1522-2594) 72 (6), 1712–1722. https://doi.org/10.1002/mrm.25083.
Tobisch, A., Stirnberg, R., Harms, R.L., Schultz, T., Roebroeck, A., Breteler, M.M., Stöcker, T., 2018. Compressed
sensing diffusion spectrum imaging for accelerated diffusion microstructure MRI in long-term population imag-
ing. Frontiers in Neuroscience (ISSN 1662-453X) 12, 1–19. https://doi.org/10.3389/fnins.2018.00650.
van der Kouwe, A.J.W., Benner, T., Salat, D.H., Fischl, B., 2008. Brain morphometry with multiecho MPRAGE.
NeuroImage (ISSN 1053-8119) 40 (2), 559–569. https://doi.org/10.1016/j.neuroimage.2007.12.025.
Weiskopf, N., Suckling, J., Williams, G., Correia, M.M., Inkster, B., Tait, R., Ooi, C., Bullmore, E.T., Lutti,
A., 2013. Quantitative multi-parameter mapping of R1, PD(*), MT, and R2(*) at 3T: a multi-center valida-
tion. Frontiers in Neuroscience (ISSN 1662-4548) 7 (June), 95. https://doi.org/10.3389/fnins.2013.00095. http://
www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3677134&tool=pmcentrez&rendertype=abstract.
Zaiss, M., Ehses, P., Scheffler, K., 2018. Snapshot-CEST: optimizing spiral-centric-reordered gradient echo acqui-
sition for fast and robust 3D CEST MRI at 9.4 T. NMR in Biomedicine (ISSN 1099-1492) 31 (4), 1–14. https://
doi.org/10.1002/nbm.3879.
CHAPTER
Advanced reconstruction methods for

fast MRI
2
Florian Knolla,b
a NYU
School of Medicine, Department of Radiology, New York, NY, United States
b FAU Department of Artificial Intelligence in Biomedical Engineering, Erlangen, Germany
2.1 Introduction to image reconstruction for fast MR imaging

Due to the nature of the acquisition physics and the use of magnetic gradient fields to perform spatial
signal encoding, MR data points are acquired in frequency space (Fourier-, or k-space). This Fourier
nature defines the essential properties of the image acquisition and reconstruction process. The number
of required measurement points to obtain an image without aliasing artifacts is set by the well-known
Nyquist–Shannon sampling theorem. The relationships between image domain and Fourier domain
explain both the types of image artifacts that are specific to MRI due to certain corruptions during data
acquisition, as well as particular features that can be facilitated to make the measurement process more
efficient.
As the name suggests, image reconstruction is the task of obtaining corresponding images from
measurement data that are acquired by the MR scanner. Image reconstruction is a field that has evolved
constantly since the first conception of MRI, where it was performed with what would now be called
a back-projection approach, similar to what is used in computed tomography. Developments in image
reconstruction are closely related and often driven by corresponding developments in pulse sequences,
which require new data processing strategies or correction mechanisms. Equally, new ideas for data
acquisition are often inspired by progress in image reconstruction. The goal of this chapter is to pro-
vide an overview of a selection of developments in image reconstruction for fast MR imaging. It will
cover both well-established methods that go back to the late 1980s, as well as currently ongoing re-
search efforts. Pulse sequence developments are not included in this chapter, because they are the
topic of the previous chapter. The specific topic of simultaneous multi-slice (SMS) acquisitions and
their applications to neuroimaging are also not included here, because they are covered later in Chap-
ter 3 (Simultaneous Multi-Slice MRI). Since this chapter does not cover details about pulse sequences
and spin physics, a good starting point is the abstraction level of the well-known MRI signal equa-
tion:
+∞ +∞
fj (kx , ky ) = cj (x, y)u(x, y)e−i(kx x+ky y) dxdy. (2.1)
−∞ −∞
Advances in Magnetic Resonance Technology and Applications, Volume 4, ISSN 2666-9099. https://doi.org/10.1016/B978-0-12-822479-3.00010-5
21
Another random document with
no related content on Scribd:
antiche, nè operando ancora le nuove istituzioni, una plebe iraconda,
avida, criminosa rimase despota di Parigi e della Francia; il mondo,
che alla parola di repubblica avea sperato la grande pacificazione
della democrazia, si sgomentò quando la vide, da rigeneratrice della
dignità umana, cangiarsi in sovvertitrice della società e di ciò che
l’uomo ha più sacro, la famiglia, la proprietà.
Come nel 1830, ogni paese risentì di quell’urto; e dove fin là erasi
aspirato ad acquistare o realizzare il Governo costituzionale, si prese
ad abbatterlo; il rinascimento italiano da difensivo si mutò in
aggressivo. Le potenze straniere aveano dato mano al movimento
pacifico, esortato i popoli a fidare ne’ principi, promesso a questi non
solo l’appoggio morale della parità d’istituzioni, ma anche il
materiale, caso mai l’Austria attraversasse il quieto decorso.
L’importanza consisteva dunque nel non turbare la pace: quando
l’Austria la turbò coll’occupare Ferrara, trovossi vinta e costretta a
recedere: guaj al momento che le fosse ridonata la superiorità col
prendere noi l’offensiva!
Ma la Francia repubblicana come intenderebbe i suoi politici doveri?
Lamartine, che colla poetica frase avea fatto aggradire il Governo
repubblicano, comparve eroe nel sostenerlo contro il furore plebeo;
ma costretto a condiscendere a tutti, e adulare come ogni potere
nuovo, e sprovvisto di tutt’altra idea che quella dell’opposizione,
trovavasi incapace di riordinare, e di concepire un avvenire altrimenti
che fantastico. Qual ministro degli affari esteri, all’Europa dichiarò (2
marzo) che, a differenza di quella del 1793, la repubblica non
minacciava ai Governi, comunque fossero costituiti, conoscendo
pericolosa alla libertà la guerra; considerare i trattati del 1815 come
non più esistenti, ma rispettare le circoscrizioni territoriali stabilite in
essi; se però qualche nazionalità oppressa si svegliasse, «se gli
Stati indipendenti d’Italia fossero invasi, od impacciate le interne loro
trasformazioni, Francia tutelerebbe i legittimi progressi».
Dire quanto basti per sospingere i passionati, e intanto riservarsi
pretesti onde rinnegarli; sopreccitare ne’ popoli l’amore della libertà e
indipendenza, eppure assodare i trattati che le conculcavano;
estendere la periferia morale, e impedire la materiale trincerandosi
nell’amor della pace, era indegno d’una gran nazione. Vero è che il
Gioberti, testimonio degli errori parigini, scriveva qua lettere della
sua solita esagerazione contro la repubblica; i festeggiamenti fattine
a Roma malgrado il papa, indicavano la mano di un Buonaparte che
ne sperava profitto; i poveri, attruppatisi a Napoli per chieder lavoro,
a Firenze per non anticipare le pigioni, a Genova per partecipare ai
guadagni de’ negozianti, palesavano una feccia che presto al
fermento sormonterebbe. Ma i popoli restano sordi agli avvertimenti
per non badare che alle catastrofi; e inebriati da quell’esempio, e
illudendosi su quelle parole, credettero mature le sospirate
franchigie.
Se il desiderio d’italianità nella restante penisola esprimevasi in
applausi ai regnanti (1847), nel Lombardo-Veneto concentravasi in
fremiti. Delle riforme amministrative concedute ai vicini già era in
possesso da gran tempo questo paese, mercè l’antica tradizione
municipale; nè qui si cercava rigenerare, bensì distruggere il
Governo: scopo determinatissimo, proponendo l’acquisto di quella
nazionalità, senza cui non parea possibile libertà soda, potente
dignità, verace progresso. Ma non se ne vedea modo che in un
subbuglio europeo. Aspettando il quale, la folla coglieva ogni destro
di esprimere avversione ai dominanti, simpatia ai principi italiani, un
accordo di volontà, ben diverso dalle congiure, qui men che altrove
opportune dove lo scopo era palese, e robustissima la
repressione [61]. Gaetano Gaisruk, arcivescovo di Milano, era
vilipeso per iscarsa dottrina e ignoranza di ecclesiastiche discipline,
ed esoso come straniero fin quando la sua morte fece rendere
giustizia a una generosissima beneficenza, a un sentimento di
giustizia che non lasciavasi raggirare dai circostanti, nè da influssi
d’anticamera, di sacristia, di consorzj; alla sua franchezza di esporre
rimproveri ai subalterni e ragioni ai potenti; alla cura degl’interessi
generali di questo paese che forse non amava, e da cui non era
amato. Vero è che non sapeva di lingua e zoppicava di stile.
Ai funerali di lui proruppero il vilipendio vulgare e poetici insulti; poi si
spiegò così solennemente il voto d’avere un prelato italiano, che
l’Austria vi destinò il bergamasco Romilli. Nè le virtù, nè il sapere, nè
l’attitudine e la prudenza di lui erano conosciuti: che importava? egli
era italiano, e bastò perchè, come a Pio IX, così a lui si facessero
feste strepitose (5 7bre), con iscrizioni allusive a patria, a Italia. Ma i
Viva non furono accompagnati dai soliti Mora; la turba,
dall’applaudire al palazzo arcivescovile, passò al fischiare sotto le
finestre ove agonizzava uno degli uomini più splendidamente
benefici [62]; poi agli inni a Pio IX seguirono i disordini che
riscontrammo in ogni altro luogo, e come in ogni altro luogo i
poliziotti dovettero tirar le sciabole: prima volta che la turba milanese
affrontasse la forza, prendendola in disprezzo perchè la sua
moderazione credette impotenza.
Poco appresso adunavasi il settimo Congresso scientifico a Venezia;
e sebbene vi mancassero Piemontesi, Toscani, Romagnoli, atteso
che già possedevano quello a cui i Congressi erano avviamento,
parve injettare la vita in quella città man mano che procedeva: le
discussioni scientifiche ed economiche assunsero importanza
politica; la quistione delle strade di ferro, che già avea agitato
Genova, qui fu colta con tale aspettativa, che a pena agli ascoltanti
bastò la sala del gran Consiglio: la quale poi nell’adunanza finale,
cogli applausi dati a qualche scienziato, e negati al vicerè, vide
prorompere manifestamente la volontà paesana.
Sentì il vicerè l’insulto, e ne fece cadere la vendetta sopra
l’applaudito: ma che ivi si concertassero i capipopolo per iniziare la
rivoluzione, è falso [63]. Nè società segrete o comitati direttori
promossero le dimostrazioni, che da quel punto si moltiplicarono in
tutte le città (1848). La più significativa fu l’astenersi dal fumare:
sucida abitudine venuta qua d’oltr’Alpi, e il cui abbandono poteva
esprimere e un ritorno all’urbanità, e che la gioventù possedeva
volontà unanime, e conosceva la forza dell’abnegazione; due qualità
indispensabili al risorgimento nazionale.
L’astinenza volle spingersi fino a violentare altri; e sia vero o no che i
militari o la Polizia mandassero attorno fumanti provocatori, ne
nacque occasione (5 genn.) di trarre le sciabole; il popolo fu ferito e
calpesto, come sempre, e come già in tutti gli altri paesi d’Italia; il
numero delle vittime fu esagerato, ma compiante per tutta Italia quai
martiri; le declamazioni de’ circoli e de’ giornali e le esequie
drammaticamente ripetute in ogni angolo affondavano sempre più
l’abisso tra noi e gli stranieri.
La Congregazione Centrale, corpo che rappresentava il paese e che
non aveva sino allora conosciuto altro dovere che di eseguire la
volontà superiore, sentì pur quello, impostogli dalla propria
istituzione, d’ammonire il potere, d’iniziare miglioramenti, di
presentare rimostranze. Il bergamasco Nazzari ne sporse una, dove
non chiedeva innovamenti, ma l’attuazione della sovrana ordinanza:
che se altre in altri tempi l’autorità aveva lasciate cascare, l’aura
odierna impose che la petizione fosse accolta, appoggiata, spedita a
Vienna. Allora la paura dell’opinione pubblica assunse la maschera
di coraggio civile; le Congregazioni provinciali e le municipali e le
Camere di commercio presentarono istanze e richiami, esitanti fra il
rispetto abituale e una risolutezza insolita: pure restringevansi
prudentemente a chiedere si mettesse in atto ciò che già era in
decreto, o a trarne le legittime conseguenze. Anche gli scritti di
qualcuno che esponeva per la stampa estera la condizione e i
bisogni del paese, non parlavano che delle riforme necessarie per
riconciliare la provincia coi dominatori, e far meno indecorosa la
servitù.
Pari agitazione legale nel Veneto; e citando leggi inosservate, si
domandò una censura meno assurda, e di partecipare al decidere
sugli interessi immediati del paese; insomma che, rientrando nelle
vie della morale e della civiltà, si togliesse l’onnipresenza deleterica
della Polizia, odiata più veramente che non il Governo.
Secondare questo movimento legale sarebbe stato il modo di
calmarlo sinceramente, o fintamente eluderlo; ma il vicerè
conosceva solo arti diverse: il Nazzari esprime i voti della
rappresentanza paesana, ed esso ordina sia sorvegliato dalla
Polizia: Manin e Tommaseo espongono domande a Venezia
(gennajo), ed esso li fa arrestare: crescendo l’irrequietudine di
Milano, promette chiedere ampj poteri da Vienna per soddisfarvi, e
gli ottiene, e bandisce si rassicurino perchè omai egli si recherà in
mano le redini dello Stato; e la notte stessa manda ad arrestare
persone, diversissime d’indole, di relazioni, di costume, e senza pure
una parola, deportarle in Germania. Contemporaneamente fecero
dal mitissimo imperatore dichiarare (22 febb.), lui avere operato
abbastanza pei popoli, nè essere disposto a ulteriori
condiscendenze; affidarsi nel valore delle sue truppe; e gli chiesero
l’arbitrio d’arrestare, di deportare, di bandire la legge marziale.
Questi atti tolsero ogni confidenza nel Governo, che non trovando
chi cospirasse, perseguitava, eppure tremava davanti a un popolo
che irritato, non sbigottito, opponeva il silenzio e l’astinenza.
Un potere minacciato diviene violento; parlavasi di truppe sempre
nuove giungenti in Italia, di promessi saccheggi, di bombardamenti
al minimo agitarsi. E per verità, risoluti com’erano a reprimere colla
forza, sarebbero dovuti porsene in grado [64], dacchè fiutavasi in aria
la rivoluzione a segno, che Metternich ripeteva a tutti gl’incaricati
d’affari, — Sta primavera in Italia vi avrà bôtte e ferite»; poi il vicerè
partiva, lasciando la legge stataria come suo legato a un paese
dov’era seduto vent’anni. D’altro lato susurravasi d’armi ammassate
in Milano, di corpi predisposti dai profughi ai confini, di incitamenti
uffiziali venuti dalla Francia, dall’Inghilterra, più dal Piemonte: eppure
il successo chiarì che nè armi v’avea, nè intelligenze, nè preparativi;
gli stessi Mazziniani aveano di quei giorni a Parigi preso accordo di
non alterare colle loro mosse il quieto svolgimento italiano, e la
Giovane Italia erasi adagiata nelle braccia di un’Associazione
nazionale. Il martirio si venera, ma non si predica: e quale
onest’uomo assumerebbe la responsabilità d’avventare il paese
inerme nel terribile esperimento d’un’insurrezione contro un esercito
sì bene disciplinato? Pure la pazienza cessa quando cessa la
speranza, e giunge un’ora in cui per le nazioni l’obbligo della fedeltà
cede al diritto d’acquistare la sicurezza che più non trovano
nell’ordine stabilito; e quest’ora la Provvidenza la batte
ineluttabilmente. E come i colpi provvidenziali scoccò d’onde meno
sarebbesi aspettato.
Vienna, città che credevasi ridotta materiale nei godimenti, e
particolarmente divota ad una dinastia che la faceva capo di un
grande impero, erasi stancata dello stupefacente assolutismo di un
ministro, che facendo sinonimi governare e comprimere,
catalogando gli uomini secondo quel che pagavano, riducendo il
Governo a doganieri, burocratici, spie e soldati, privavalo della sua
più nobile qualità, l’iniziativa; dei sudditi spodestava le volontà, e
scroccando il nome di accorto e robusto coll’impedire ogni
movimento, lasciossi sopraggiungere da uno di que’ turbini, che cogli
abusi svelgono anche le istituzioni. Ambiziosi di palazzo e di
gabinetto secondarono gli aliti liberali, già incitati dalle diatribe della
Germania settentrionale, poi dai movimenti slavi e dalla rivoluzione
francese: la Boemia e la Gallizia avevano mandato a chiedere libertà
di stampa, d’insegnamento e d’altro: un proclama dell’ungherese
Kossuth allora allora divulgato, ove si chiedeva che l’impero si
riformasse, alle singole nazionalità il governarsi, e congiungendole in
federazione, assegnò più preciso scopo alle domande della Società
Politica e della Industriale di Vienna, e degli studenti che inanimati
dalla sollevazione di quelli di Monaco, proclamarono una petizione
nell’aula universitaria, e vollero portarla alla Corte (13 marzo).
Questa oppose il niego, poi i sopratieni, ma il popolo tumultuava; gli
eserciti stavano lontani; la piccola guarnigione poteva esser presa in
mezzo dagl’insorgenti: i quali, inviperiti da alcuni colpi da essa tirati,
mostrarono inaspettato coraggio e impetuosa fermezza; e mentre i
ministri e la Corte vacillavano in quell’inaspettatissimo accidente, si
ottenne fosse espulso Metternich, e per tutto l’Impero libera la
stampa, guardia nazionale, un’assemblea generale per formare la
costituzione; e il buon Ferdinando proclamava: — Seriamente,
solennemente, e con sincera soddisfazione andai incontro ai voti del
mio popolo, concedendo una costituzione, ch’io riguardo come l’atto
più soddisfacente della mia vita». Applausi, abbracci, inni
festeggiano l’affratellamento; i liberali esultano del loro trionfo, e
frenano la plebaglia ladra; e la Corte, affidando il ministero a
Pillersdorf e ad altri onesti della vecchia scuola, spera pure col
tempo rivalere contro le esigenze superlative.
Il telegrafo portò in Lombardia (17 marzo) quelle concessioni
viennesi; e la loro dissonanza dai minacciosi rifiuti dei giorni
precedenti vi dava l’aria d’un’inevitabile necessità; l’Austria doversi
trovare agli estremi se mettevasi per una via da lei esecrata, e su cui
non era possibile durasse. Pertanto alle fantasie già bollenti s’offre
l’incentivo dell’occasione: preceduti dalla rappresentanza
municipale, i Milanesi vanno a domandare armi per la guardia civica;
e ne hanno la promessa, fra i Viva e le coccarde; ma quando
convengono al palazzo municipale per riceverle, eccoli assaliti dalla
truppa, che alla ventura ne coglie alquanti, e li trascina in fortezza.
L’indignazione precipita il moto, che già era cominciato non senza
sangue; l’esultanza si converte in furore; e mentre alcuni
persistevano a consigliare che s’accettassero le concessioni, e
consolidandole si facessero scala a maggiori, altri elevano le
speranze fino all’indipendenza; impennati i tre colori, gridano «Viva
Pio IX, e Morte ai Tedeschi»; ubriachi di magnanima imprudenza
rimettono la suprema decisione ai rischi dell’audacia; e vendicando
le paure di cui si era loro prodigato l’oltraggio, cominciano una
battaglia memorabile (1848 18 marzo). Dappertutto sbarrar le vie
con quel che prima venisse alla mano; e se mancassero le travi, le
botti, i lastroni delle vie, s’accatastano i mobili anche più fini, quasi si
sentisse bisogno di fare più costosi sagrifizj. Capita una carrozza?
ne staccano i cavalli, la rovesciano, la riempiono di ciottoli, di strame,
e il passo è intercetto. Ogni casa era munita a guisa di fortezza; sui
davanzali panieri di sassi, e dalle socchiuse gelosie sporgeano
canne mortifere, e dentro preparati coltroni e materassi per
ammortire i colpi o spegnere le bombe. Alla scarsezza di fucili e di
munizioni supplivasi come si poteva, ammannivasi cotone
fulminante, spogliavansi i musei d’armi. I nemici entro le caserme e
dal duomo si difendeano; aprivansi la via sanguinosamente, traverso
una tempesta di tegoli e di ciottoli, per riunirsi attorno al castello,
dove accampavano sotto una pioggia, incessante come il tempellare
delle campane, che mentre infondeano terrore nel nemico,
incoravano gl’insorgenti dando certezza ai lontani che quella chiesa,
quel quartiere erano sgombri. Alcune vie furono prese e riprese; e si
sparse e si credette che i Croati si piacessero di gratuite e raffinate
atrocità, sventrare incinte, crocifiggere od arrostire a lento fuoco i
vecchi, spiaccicar fanciullini, o infilzati portarli sulle bajonette; altri
sepellire vivi, o coprire d’acquaragia e poi infiammare. Quando poi
leggemmo su altri giornali apposte le medesime spietatezze ai nostri
contro i Tedeschi, comprendemmo che è stile delle nazioni odiantisi
il ricambiarsi tali accuse. Certamente abbondarono atti e di ferocia e
di magnanimità; e gran coraggio vi volle perchè con pochi fucili da
caccia, gente da studj, da officine, da bottega per cinque giornate
tenesse fronte a truppe disciplinate. Nè le armi che vantavansi
apparecchiate, nè i fuorusciti o i Piemontesi o i campagnuoli che
diceansi aspettar solo un cenno, comparvero; sebbene per via di
palloni areostatici si diffondessero appelli e incoraggiamenti. Ma
neppure il nemico era allestito a difesa; e le insufficienti e deteriorate
sue munizioni, la concorde perseveranza de’ cittadini, il probabile
dilatarsi dell’insurrezione nella campagna, l’incertezza di ciò che
accadeva a Vienna, l’apprensione che i Piemontesi arrivassero,
indussero il maresciallo Radetzky (22 marzo) a ordinare la ritirata. E
Milano si trovò libera, con un’esultanza più viva quanto meno
aspettata, compra con trecencinquanta vite, fra cui quaranta donne e
trentaquattro fanciulli.
Scene simili eransi rinnovate in altre città. A Como uscirono subito
ajuti di rifuggiti dalla Svizzera, e con ostinata battaglia per le vie
costrinsero i Croati a capitolare. Il lago, il Varesotto, la Brianza
disarmano o cacciano gli stranieri, mandano prodi a soccorrere
Monza e Milano: la Valtellina con poca fatica si libera anch’essa, le
scarse truppe lasciando ritirarsi in Tirolo. A Bergamo un cappuccino
col Cristo e la bandiera italiana chiama il popolo alla libertà, e a capo
di risoluti move ad ajutar Milano; mentre in città erano prese le
caserme e l’arciduca Sigismondo, al quale o generosità o abitudine
servile concesse di ritirarsi, come pure ai Croati. A Brescia lasciasi
passare il fuggiasco Raineri, ma si getta un petardo ai Gesuiti; poi
appena proclamate le concessioni, il generale Schwarzenberg
scorre la città applaudito: il reggimento Haugwitz ivi acquartierato
era quasi tutto d’Italiani; onde credendo l’impresa finita, non si corse
ad ajutar Milano, e si lasciò passare senza ostacolo l’arciduca
Sigismondo, fuggente da Bergamo. I paesi della Franciacorta, della
Riviera, delle Valli insorgono, e tutto è libero fino al Tirolo. Allora i
Bresciani, accorti del vero, intimano a Schwarzenberg di cedere, e
poichè resiste, cominciano la lotta, trucidano il suo ajutante
Hohenlohe che veniva a esibir pace, e a gran fatica il generale
stesso si sottrae; lasciossi partire con onorevole capitolazione e
coll’armi la truppa, la quale postasi agli Orzi sull’Oglio, potè
spalleggiare la ritirata di Radetzky. Questo, nottetempo staccatosi da
Milano per porta Romana, a Melegnano incontrò qualche tentativo di
resistenza, ma colla severità sbigottì a segno, che nessuno più gli si
oppose su tutta la via, dove ogni pianta, ogni rivo, ogni ponte potea
divenire un ostacolo funestissimo. Solo dopo passato l’esercito si
gridava libera Lodi. In Cremona un reggimento d’Italiani fraternizzò
cogl’insorgenti; sicchè il generale Schönhals capitolato partiva con
quattrocento ulani e la cassa e le armi, lasciando alla città due
battaglioni di fanti, una batteria da campagna. A Pizzighettone fu
presa la fortezza con diciotto cannoni e settecento casse di
munizioni, che furono trasferiti a Cremona, invece di raccorre colà
anche gli altri e chiudere il passo dell’Adda, o ingrossare al ponte di
Lodi e assalire Mantova.
L’occupazione di questa fortezza sarebbe stata decisiva dei casi
nostri; e Gorczkowsky che la comandava, seppe trastullare i cittadini
colla guardia civica, in modo che non pretendessero la cittadella:
intanto i savj e i vescovi raccomandavano la quiete, per timore che la
fortezza fulminasse la città. Ed ecco giungere un indirizzo del
municipio di Trento, esprimente il proposito di staccarsi dal Tirolo per
far causa comune coll’Italia, esibendole persone e averi. Vi si
risposero parole; si lasciò passare il duca di Modena; si accolsero
soldati in ritirata, i quali presto furono bastanti ad assicurare la città
agli Austriaci. Visto l’errore, si gridò tradimento quel ch’era stato
difetto di sagacia e di coraggio. Dappertutto le Congregazioni
municipali e l’alto clero aveano procurato rattenere da atti, dai quali
non poteva ripromettersi altro che ruina; dappertutto fu risparmiato
l’inutile sangue, contro la dominazione, protestando solo colla gioja
del liberarsene.
Venezia, scarcerati Tommaseo e Manin (17 marzo), li portò in trionfo,
al proclamarsi la costituzione e la libertà della stampa,
rimbombarono i Viva all’imperatore; ma l’annunzio della insurrezione
di Milano fece comprendere altre possibilità, e i civili stettero contro
la forza. Venezia poteva essere bombardata dall’arsenale e dalla
goletta del porto; ma Palfy governatore si peritò, nell’incertezza di
quanto accadeva a Vienna, e alla magistratura municipale concesse
d’armare la guardia civica. Intanto bucinavasi di tradimenti orditi dal
nemico, e che Merinovich, odiato comandante all’arsenale,
preparasse materie da incendio, quando i suoi dipendenti gli si
avventarono e l’uccisero (22 marzo): l’avvocato Manin, postosi a
capo de’ cittadini, tra la persuasione e la forza occupa l’arsenale; il
governatore rassegna i suoi poteri a Zichy comandante militare, e
questo fa colla municipalità una capitolazione, per cui possa menare
via la truppa tedesca, con tre mesi di paga, lasciando la cassa, le
armi, i soldati italiani a Venezia. Tredici persone furono spente: ai
nemici nessun insulto; anzi la generosità arrivò a tale imprudenza,
che volendosi mandare a Pola l’ordine alla flotta di venire
all’obbedienza degli insorgenti, si affidò l’avviso al legno stesso che
portava Palfy a Trieste. In conseguenza questo potè prevenirla, e
Venezia restò paralizzata del suo braccio destro, la flotta.
Però essa trovavasi libera legalmente; e il popolo espose la
Madonna di San Marco, come poi fece in ogni gaudio e in ogni
sventura: si elesse un Governo provvisorio (23 marzo) con Castelli,
Tommaseo, Paleocapa, Camarata, Pincherle, Solera, Paolucci,
Toffoli, e a capo Manin, e si proclamò la repubblica, estesa allora
nulla più che la piazza San Marco. Ma le città di terraferma non
tardarono ad imitarla, cacciando o disarmando i soldati; il generale
d’Aspre è costretto abbandonar Padova; il forte di Malghera viene
occupato dalle guardie civiche di Mestre, quello di San Felice dai
Chiozzotti; quelli di Osopo e di Palmanova si arrendono, e n’è posto
comandante il generale Zucchi, che dal 1831 vi rimanea prigioniero.
A Verona stava il vicerè, il quale colle promesse tenne a bada i
cittadini, e salvò così il nido dove l’aquila rinnoverebbe le penne.
Tutte le città si diedero Governi proprj, che poi si fusero nel
veneziano. L’esercito austriaco in quei giorni perdè quattromila morti,
settemila prigioni e feriti, diecimila prigionieri, oltre i settemila di
Venezia.
Anche in Modena si leva rumore, e il duca, istituita una giunta, si
ritira sul territorio austriaco, mentre il granduca occupa i territorj di
Massa e Carrara. Il duca di Parma (10 marzo), udito la sollevazione
di queste città ove combattendo i militari tedeschi, cinque cittadini
ebbero morte e molti ferite, ma costrinsero i nemici a deporre le
armi, non solo si rammorbidisce come tutti gli altri, e promette lo
statuto, ma deplora d’aver subito l’influenza straniera, e dichiara
rimettere i suoi destini a Pio IX, Carlalberto e Leopoldo, perchè
facciano de’ suoi Stati quel che meglio comple all’Italia, pronto a
ricevere egli quel compenso che crederanno conveniente; ed egli se
n’andò in Romagna, suo figlio a Milano per offrirsi alla causa italiana,
dove invece fu tenuto prigione.
L’insurrezione di Milano erasi sentita dai Piemontesi (19 marzo) con
tutto l’interesse di nazione e di vicinanza; e l’intera popolazione
fremea perchè si corresse a sottrarre la vicina da uno sterminio
inevitabile; già molti vi si spingeano volontarj, malgrado le guardie
poste al confine, e vi si mandavano munizioni. Poco prima,
Carlalberto, risoluto di mettersi francamente nelle norme
costituzionali, aveva chiamati al ministero Sclopis, Franzini,
Boncompagni, Desambrois, Revel e i genovesi Pareto e Ricci, sotto
la presidenza di Cesare Balbo. La costoro popolarità, le conosciute
intenzioni, i voti gridati, anzi intimati a loro dai Genovesi, li faceano
scopo a smisurate speranze. E poichè in capo d’ogni speranza stava
l’italianità, tutti chiedevansi se il Piemonte trarrebbe la spada per
rivendicarla. Non era questo il lungo voto di Carlalberto? non teneva
egli in piedi settantamila armati, e riboccanti gli arsenali, e pingue il
tesoro, e uno stato-maggiore incomparabile, e tutta l’uffizialità
anelante di provarsi cogli oppressori?
Le realtà stavano a gran pezza dai discorsi. Il preconizzato sistema
militare del Piemonte appariva disadatto a trasformarsi
subitaneamente dal piede di pace in quello di guerra attiva; artiglieria
e cavalleria eccellenti ma scarse; le riserve male esercitate, e
avvezze al riposo e agli affetti domestici; i soldati coraggiosi
personalmente, ma non altrettanto disciplinati tutti insieme; uno
stato-maggiore più di comparsa che di valentìa; nessuno poi avea
mai fiutato battaglie; nè in quel precipizio più di dodici in quindicimila
uomini si potrebbero mettere in campo; e di questi un buon dato
eransi spediti in Savoja per impedire un’irruzione dei Voraci, bande
comuniste della Francia. Dell’Austria ignoravasi lo sfasciamento;
poco si poteva ripromettersi dalla restante Italia, inavvezza all’armi;
l’Inghilterra, che a consigliare e moderare l’italico movimento avea
spedito lord Minto, non che attizzasse come si spargea, dichiarava
essere la Lombardia assicurata all’Austria dai trattati medesimi che
assicuravano Genova al Piemonte, e il toccar l’una
comprometterebbe l’altra. I soccorsi della Francia metteano ribrezzo
or ch’era repubblicana, potendo divenir rovinosi al principato; e il
famoso motto attribuito a Carlalberto Italia farà da sè era una
protesta contro quegli ajuti sgraditi. D’altra parte i veggenti, persuasi
che si consolidano più cause coi temperamenti della prudenza, che
non se ne guadagnino colla furia, aveano sempre sconsigliato il
Piemonte dalla guerra [65]; ai nuovi ministri era riuscito di
consolazione l’accertarsi che l’Austria non minacciasse il Piemonte,
il quale potrebbe tranquillamente assodare, svolgere, applicare la
donatagli libertà. E in fatti il programma loro esprimeva: fare
preparativi se mai l’Austria chiarisse guerra, ma non provocarla:
riconoscere la Repubblica francese; allearsi coll’Inghilterra e cogli
Stati costituzionali d’Italia purchè non rompessero a ostilità.
Carlalberto, sempre fisso ad un fine, tentennava sui mezzi e sul
tempo, e viepiù da che si sentì trascendere dal movimento. — Che si
dice sottovoce al Congresso di Genova?» interrogava. — Si dice,
Viva Carlalberto», gli si rispondeva. Ed egli: — Ma più basso si dice
Viva Mazzini». In una delle più solenni festività di quel festivissimo
tempo, tutte le comunità del regno vennero a solennizzare (25
febbr.) la promessa costituzione, e sfilarono tripudianti di bandiere, di
inni, di Viva innanzi al re, e soli mesti e abbruniti noi Lombardi,
sfuggiti al carcere e alla legge marziale. Chi l’ha veduta non potrà
mai più dimenticare quella giornata, d’accordi non anco turbati, di
speranze potenti di tutto il prestigio, d’una libertà di cui nessuno
erasi disamorato. Sarebbe stata la più bella della vita di Carlalberto;
ma la sera giunse l’avviso della repubblica proclamata a Parigi, e noi
gli udimmo dire: — Anche questa vicenda farà il giro d’Europa. Poco
mi cale di me: duolmi de’ miei figliuoli; ma non importa purchè il mio
popolo sia felice».
Proposizioni a lui erano state rivolte da Lombardi prima della
sollevazione; ma non le ascoltò egli direttamente, bensì un suo
ajutante: pure, in iscritto confidenziale, ripetè la promessa mandata
ai comizj di Casale, che, dato il caso, guiderebbe il movimento
patriotico d’Italia. In Milano i proclami animavano alla difesa colla
certezza degli ajuti piemontesi; da’ campanili speculavasi il loro
arrivare; fin Radetzky vi credette: da cittadini ricchi e reputati si
sottoscrisse un invito a Carlalberto perchè soccorresse e prendesse
la Lombardia (20 marzo); eppure Carlalberto che l’avea chiesto,
esitava ancora, e i ministri davano assicurazione di buona vicinanza
all’ambasciadore austriaco. Ma la gioventù freme guerra; i portici di
strada Po e la piazza della reggia formicolano di gridanti guerra;
guerra vuole l’Università: e quelli che non sanno figurarsi la libertà se
non a cavallo d’un cannone. Il re e i ministri sapeano che perde
l’autorità chi la sottopone al tumulto: ma e se Milano soccombesse a
un nuovo Uraja? qual onta pel vicino armato? E che farebbe
Genova, la quale avea gridato Con Milano, se no, no? la
compassione non potrebbe prorompere contro il principe, e fino a
gridare la repubblica?
Mentre vacillavasi tra i consigli della prudenza ed i precipizj della
generosità, ecco giunge (22 marzo) che Milano s’è liberata da sè;
che i Tedeschi rotti e scompigliati vanno in pienissima fuga fra le
strade rotte e le campagne inondate, incalzati dalle popolazioni,
risolute a non lasciarne vivo uno, uno solo [66].
Carlalberto gettò la propria spada sulla bilancia dei ministri, e
proclamò che coi suoi proprj figli si metteva a capo dell’esercito,
portando alla Lombardia «i soccorsi di fratello a fratelli; di guiderdone
non si parli: solo a guerra finita si deciderà delle sorti del paese».
Ammirazione, gioja, affetti si rovesciano allora sopra Carlalberto, il
migliore, il più grande dei re, la spada d’Italia; se ne dimenticano i
torti, prima ch’egli dichiari dimenticati quelli de’ sudditi; egli si
rassegna a venir ricevere sul balcone e per le strade le acclamazioni
da cui sempre aveva aborrito; assiste al Tedeum cantato
dall’arcivescovo di Torino, a cui quest’atto non risparmia i fischi;
passa in rassegna la plaudente guardia nazionale, contento che sui
vecchi suoi giorni rifulga quel raggio di speranza, che aveva indorato
i vigorosi.
Gli altri paesi d’Italia rispondono a quel grido. A Roma Ciciruacchio
mena la folla ad abbattere lo stemma del palazzo d’Austria, e
occuparlo a nome della Dieta italiana, della quale s’intima a Pio IX di
farsi capo, mentre le campane suonano, i cannoni bombano, il Masi
improvvisa, il gigantesco padre Gavazzi bolognese predica, il
marchese Patrizj, il principe Ruspoli offrono denaro, i figli, la persona
alla causa comune: e Pio IX riconoscendo la mano del Signore in
quella vittoria (30 marzo), rammemora che «d’ogni stabilità e
prosperità è ragion prima la concordia, e che la giustizia sola edifica,
mentre le passioni distruggono»; Leopoldo granduca intuona: —
L’ora del risorgimento d’Italia è giunta improvvisa, nè chi ama questa
patria comune può ricusarle soccorso. Figli d’Italia, eredi della gloria
militare degli avi, non devono i Toscani rimanere in ozio vergognoso,
mentre la santa causa dell’indipendenza si decide, ma volare al
soccorso de’ fratelli lombardi». Il Ministero napoletano che aveva
cercato tenersi saldo contro le dimostrazioni di piazza, fu da queste
scomposto; si dovette promettere la guerra santa, capitanata da
Pepe, esule da ventisette anni, e un Ministero preseduto da Carlo
Troya, esule della stessa causa (aprile); e il re proclamava: — Le
sorti della comune patria vanno a decidersi nei piani della
Lombardia; ed ogni principe e popolo è in debito di accorrere a parte
della lotta che deve assicurare l’indipendenza, la libertà, la gloria.
Noi intendiamo concorrervi con tutte le nostre forze di terra e di
mare, cogli arsenali, coi tesori della nazione; unione, abnegazione,
fermezza, e l’indipendenza della nostra bellissima Italia sarà
conseguita; e ventiquattro milioni d’italiani avranno una patria
potente, un comune ricchissimo patrimonio di gloria, e una
nazionalità rispettata».
Tanto accordo di principi e di popoli che forti di risolutezza, invigoriti
di lunghi patimenti anelano alla virile gioja delle battaglie, acciocchè
l’Italia sia, non trofeo di altrui vittorie, ma redenta pel braccio dei
proprj figliuoli; tutti dimenticando le antiche superbie e gli antichi
rancori, e contando soltanto sulla fermezza del proposito, la
temperanza delle passioni, la concordia delle volontà, i miracoli
dell’entusiasmo.
CAPITOLO CXCII.
Guerra santa. Conquassi.
La vittoria era assai meno facile che il trionfo. Sulle orme del nemico
fuggente si cacciarono alquanti, di coraggio risoluto e intelligente; e
deh come pareano belli que’ giovani, che alfine avevano qualcosa
da fare! come ne’ loro atti sfavillava eroico, incitato, romanzesco il
sentimento! Altrettanto deforme e scomposto era l’esercito austriaco;
lacero, tutto mota e sangue, famelico, con impotente anelito di
vendetta, e temendo da ogni siepe un assalto, sotto ogni ponte una
mina, in ogni villaggio barricate e tegoli; che se davanti a quello,
scompigliato da tante diserzioni, dall’insolita guerra delle vie, dalla
privazione di riposo, dall’incertezza degli avvenimenti viennesi, si
fossero abbattute le piante, recise le vie, diffuse acque, lanciata la
morte, qual ritornava di là dai monti? Ma Radetzky ebbe ad
avvedersi ben presto che il popolo non prendeva parte a
quell’insurrezione; i campagnuoli non secondarono l’impulso delle
città, nè la bassa rispose alla risolutezza dell’alta Lombardia; sicchè
egli, neppure mai attaccato, potè giungere al Mincio, e dentro al
formidabile quadro, formato dai monti, dal mare, dall’Adige colle
fortezze di Verona e Legnago, dal Mincio con quelle di Peschiera e
Mantova, rincorare le truppe, attenderne di nuove, e coi migliori
uffiziali allestire la difesa e la riscossa.
Nè alla Potenza austriaca restava allora altro appoggio che
quell’esercito e quel capitano, il quale non lasciò di tenersi per
guardie i granatieri italiani; mancante del denaro fin per vivere due
giorni, pure affacciavasi al balcone a ricevere anch’egli applausi dal
vulgo, cui buttava il poco resto de’ suoi quattrini.
L’esercito piemontese si trovò scarso oltre ogni aspettazione e
impreparato: i generali confessavano la propria inettitudine, e
consigliavano a cercare un maresciallo ai Francesi [67]; ma questi
erano sospetti a Carlalberto. Cuore intrepido con incerto consiglio,
mancante di quell’attitudine impassibile del comando, che impone
alle fantasie popolari e affascina le volontà col supporre nel
comandante una profonda persuasione; perchè era spada d’Italia
egli credette essere la mano che bastasse a maneggiarla, e ripetè
l’ambiziosa parola Italia farà da sè, la quale [68], d’effetto drammatico
in bocca di letterati e di preti, acquistava tremenda importanza
ripetuta da un re che montava a cavallo per darvi realtà.
L’esercito arrivò tardi (29 marzo), ed anzichè precipitarsi su
Mantova, mal presidiata, e con cittadini disposti alla rivolta, entrò per
Milano e Pavia, e a marcie regolari spintosi al Mincio, valore mostrò
(8 e 9 aprile) ai ponti di Goito, Valleggio, Monzambano. Passato il
fiume, coll’inutile assedio di Peschiera s’intepidì l’entusiasmo,
aspettando il parco che arrivò solo il 15 maggio: e lungo l’Adige
distesa una linea di trentasei miglia, cominciossi una guerra di
posizioni. Ben presto sessantamila uomini si trovò Carlalberto. Vi si
aggiunsero cinquemila Toscani fra volontarj e di ordinanza;
diciassettemila Romani avvicinavansi al Po; e quattordicimila
Napoletani, oltre innumerevoli volontarj; tremila Parmigiani e
Modenesi stavano sul Mincio; cinquemila Lombardi verso il Tirolo;
bande di Veneti alle alpi Carniche.
Anfossi, Longhena, Griffini, Manara, Arcioni, Simonetta, Sorresi,
Bonfanti, Tololti, Sedabondi, Torres.... capitanavano bande; bande di
volontarj polacchi ci erano menate dal gran poeta e mistico
Mickiewitz; napoletani dalla principessa Belgiojoso, siciliani da La
Masa, altri dal belgio Thamberg, altri ancora dall’attore Modena, la
cui moglie ne portava la bandiera, e di più serj dal famoso Garibaldi;
nè mancavano preti, e l’eloquente Bassi barnabita, che nel 1836
avea tanto giovato a Palermo durante il cholera, e il padre Gavazzi
parevano santificare la causa e meritarle il nome di crociata; i
seminaristi medesimi si organizzarono per le armi: nobili impeti, a cui
mancavano la disciplina e l’unione, che sole possono dare la vittoria.
Ma improvvida fiducia in noi e improvvido disprezzo pel nemico
fecero che, quando ognuno avrebbe dovuto offrire sin l’ultimo soldo
e l’ultima stilla di sangue pel riscatto nazionale, si stiticasse sui
sacrifizj, e si dissentisse sui mezzi. Come i Lombardi eransi lusingati
di vincere democraticamente in tempo che ogni forza sta
concentrata ne’ Governi, così i Piemontesi opponevano battaglia di
fronte a un esercito di mirabile disciplina ed esperienza; mentre alla
vittoria, unico scopo, sarebbe dovuto dirigersi l’impeto nazionale,
non si seppe o non si volle effettuare la leva a stormo; si tennero in
lieve conto i volontarj che, con ottima sentita, si portarono a difesa
dei varchi alpini, benchè si vedesse il nemico avvantaggiarsi dei
subitarj, corsi ad ajutarlo dalle scuole austriache o dalle fucine
stiriane. Da cinquantamila uomini che si trovavano in Lombardia fra i
ventotto e i trentott’anni, che aveano militato; e non furono richiamati
istantanemente sotto le armi: seimila trecento ch’erano disertati dagli
Austriaci, furono rejetti dall’onor militare, e coperti di quel sospetto
che invita a tradire: invece di innestare subito i coscritti nell’esercito
piemontese, con camerati esperti, sotto vecchi uffiziali, si volle
formare un esercito lombardo, sciupando denaro e tempo,
crescendo gli scioperi e quindi gl’intriganti, e non recando ajuto alla
gran causa. Giovani baliosi non aveano vergogna di rimanersi a
casa a pompeggiare nella guardia nazionale e nelle parate, e
poeteggiare sui giornali e nelle canzoni quel coraggio che è si facile
allorchè l’occasione è lontana.
In quelle ore procellose dove sono gli avvenimenti che impongono i
dittatori, d’ogni città presero il governo le persone che si trovarono o
che vollero mettersi in una posizione di molti pericoli e di nessun
vantaggio, e ripagata coll’impopolarità. Per accentrare la resistenza
e i comandi, il Governo provvisorio di Milano faticò a vincere le
gelosie, che sono brina ad ogni fiorire di speranze italiche, e fare che
ciascuna provincia gli mandasse un deputato. Vennero scelti non
coloro che aveano tramato o intrigato, forse neppure sperato; alcuni
anzi già bersaglio della stampa demagogica [69]: sì poco era figlia di
congiure quella sollevazione, che traeva nobiltà e forza dall’intento
comune e semplice di rivendicare la nazionalità.
Ogni Governo rivoluzionario si trova debole a fronte dei compagni di
rivolta, ed esposto ai mille rischi della inesperienza, della
precipitazione, del disordine. Il nostro poi, vacillante per inesperienza
e incoerente per gli antecedenti, neppure cercossi la sanzione
popolare, tanto facile in paese sistemato a municipj. Nei momenti
sublimi in cui l’ispirazione viene dalle moltitudini, essa irradia taluni
che, cessato quel lampo, devono ricadere nelle tenebre: e caratteri
medj, i quali usano riguardo a tutti, carezzano il bene come il male
politico, potrebbero mai condurre una rivoluzione, che vive di moto,
d’azione, d’audacia? Alla nostra, mentre era nel primo lancio,
imposero la formola delle società in riposo, conservare l’ordine; nè
tampoco si seppe governare una gente, così facile a governare
perchè così facile a illudere; quando tutto era straordinario,
operavasi come in occorrenze consuete.
I prestiti volontarj sono uno spediente che piace a leggersi ne’ vecchi
repubblicani; si piange d’una fanciulla che offre l’anello di fidanzata,
d’una vecchia che dona la tabacchiera d’argento, d’un prete che
levasi le fibbie; ma che profittano ora che le forze e il denaro sono
concentrati nei Governi. Si abolivano la gabella del sale e il testatico,
mentre col sospendere i pagamenti del Monte sconcertavansi tante
famiglie; si chiedeano le argenterie domestiche e gli spogli delle
chiese, mentre tesori poteano cavarsi annunziando la suprema
necessità del vincere.
Pronte nubi offuscarono quel rosato, di cui si colora l’alba d’ogni
rivoluzione. I sistemi corruttori pregiudicano l’avvenire col far che, al
punto di cambiarli, non si trovino persone capaci a rappresentare la
nuova età; e che i vulghi, lusingati di alleviamenti e beatitudini,
ricusino gli stenti con cui bisogna conquistarli, e lo spostamento
degl’interessi e delle abitudini. In società educate così, le qualità
negative prevalgono, e guaj a chi trascende una mediocrità, palliata
col nome d’eguaglianza! nome illustre, operosità, esaltazione di
nobili sentimenti, influenza riconosciuta divengono pericolosi e
denigrati. Se non bastava dunque il trovarci inesperti degli affari,
delle armi, della vita politica; se non bastava che Tommaseo e
Cattaneo, Gioberti e Rosmini, Cibrario e Brofferio, Carlalberto e
Berchet si fossero palleggiati insulti, che poteano mettere in disparte
ma non disfare, i generosi restavano elisi dal dispetto proprio o dal
sospetto altrui, all’istante che più n’era bisogno.
Gente irritabilissima gli scrittori! E alcuni di essi, che sulle prime
esageravano l’eroismo per incitarlo, ripigliarono presto il riso
sardonico; altri, che avevano aspirato ad essere primi, non
soffersero di rimanere secondi, e sbracciavansi a rivelare gli errori di
chi non faceva come loro, e autorizzavano le ire delle fazioni, che
sempre gridansi tradite da chi non le serve come esse vogliono.
Mentre il riuscire a cose straordinarie allucina in modo da far credere
tutto possibile, i tentativi arrischiati cacciano indietro molti spiriti
sbigottiti, compromettono ciò che esagerano, ruinando ciò che
trascendono. Fra coloro dunque che, per moda o per primeggiare,
aveano invocato la tempesta, molti sgomentaronsi al vederla
scatenata; e dagli inconditi sussulti di Francia presagendo qui pure
la ghigliottina o il comunismo, corazzavansi contro coloro che pur
seguitavano a chiamare fratelli.
Mentre tutti credeansi valevoli a proporre, nessuno volea la
responsabilità del risolvere; il popolo male obbediva a governanti,
dipintigli come spregevoli; e fra le canzoni e la proclamata fraternità
nessuno avea fiducia in nessuno. Finchè trattavasi di bruciare in
effigie Guizot o Metternich, e di mettere in caricatura Radetzky, molti
faceano l’eroe; quando si trattasse di fatti, l’inerzia, che prima si
crogiolava nell’impossibilità di affrontare il nemico, dappoi coglieva
pretesto dalla facilità della vittoria, tutto asserendo finito colla
cacciata de’ Tedeschi.
Ai nuovi reggitori accalcavansi i servidori degli antichi, che cogli
antichi non volendo cadere, chiedeano compensi di persecuzioni
non sofferte; improvvisati statisti offrivano consigli; speculavasi sulle
armi, sugli impieghi, sulla pubblicità, sulla fame; dilettanti del
mestiere di delatore e di carceriero continuavano a vedere
cospirazioni e delitti, e mentre sovrastava un esercito minaccioso,
eccitavano schiamazzanti paure contro spie che non si trovavano,
contro contadini che voleano soltanto far chiasso come i cittadini.

Advanced Neuro MR Techniques and Applications 1st Edition All Chapter Instant Download

Uploaded by

Copyright:

Available Formats

Advanced Neuro MR Techniques and Applications 1st Edition All Chapter Instant Download

Uploaded by

Document Information

Original Title

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

Advanced Neuro MR Techniques and Applications 1st Edition All Chapter Instant Download

Uploaded by

Copyright:

Available Formats

Full download test bank at ebookmeta.

Advanced Neuro MR Techniques and Applications 1st

For dowload this book click LINK or Button below

Download More ebooks from https://ebookmeta.com

Advanced AI Techniques and Applications in

Advanced Materials Polymers and Composites New Research

Advanced Painting Techniques 1st Edition Ruben Gonzalez

Advanced Control of Power Converters: Techniques and

Advanced Integration Techniques Third Version Edition

Features and Management of Acute and Chronic Neuro

Basic and Advanced Laboratory Techniques in

Data Science: Techniques and Intelligent Applications

In-Young Choi, PhD

Peter Jezzard, PhD

Brian Hargreaves, PhD

Greg Zaharchuk, MD, PhD

Volume 3 – Hyperpolarized Carbon-13 Magnetic Resonance Imaging and Spectroscopy – Edited by

Visit the Series webpage at https://www.elsevier.com/books/book-series/advances-in-magnetic-

Library of Congress Cataloging-in-Publication Data

British Library Cataloguing-in-Publication Data

For information on all Academic Press publications

Publisher: Mara Conner

PART 1 FAST AND ROBUST IMAGING

PART 2 CLASSICAL AND DEEP LEARNING APPROACHES TO NEURO

PART 3 DIFFUSION MRI

PART 4 PERFUSION MRI

CHAPTER 13 Contrast agent-based perfusion MRI methods . . . . . . . . . . . . . . . . . . . . . . . . 195

PART 5 FUNCTIONAL MRI

PART 6 THE BRAIN CONNECTOME

PART 7 SUSCEPTIBILITY MRI

PART 8 MAGNETIZATION TRANSFER APPROACHES

PART 9 QUANTITATIVE RELAXOMETRY AND PARAMETER

PART 10 NEUROVASCULAR IMAGING

PART 11 ADVANCED MAGNETIC RESONANCE SPECTROSCOPY

PART 12 ULTRA-HIGH FIELD NEURO MR TECHNIQUES

Karin Markenroth Bloch

In-Young Choi, Kansas City, KS, United States

Fast and robust

Recommendations for neuro MRI

1.1 MRI hardware

1.2 From signals to biomarkers

1.3 Spatial encoding strategies

1.4 Large-scale population imaging

1.5 Example multi-purpose protocols

Table 1.2 (continued)

1.6 Acquisition of neuro MRI contrasts

1.6.1 Brain anatomy

1.6.2 Tissue microstructure

1.6.3 The brain at work and rest

1.6.4 Brain perfusion

1.6.5 Biophysical tissue properties

1.7 Conclusions and future prospects

Advanced reconstruction methods for

2.1 Introduction to image reconstruction for fast MR imaging

You might also like