Functional Neuroradiology Principles and

Clinical Applications 2nd Edition Scott H. Faro

install download

https://ebookmeta.com/product/functional-neuroradiology-
principles-and-clinical-applications-2nd-edition-scott-h-faro/

Download more ebook from https://ebookmeta.com

We believe these products will be a great fit for you. Click
the link to download now, or visit ebookmeta.com
to discover even more!

Niedermeyer s electroencephalography basic principles

clinical applications and related fields Donald L.
Schomer & Fernando H. Lopes Da Silva

https://ebookmeta.com/product/niedermeyer-s-
electroencephalography-basic-principles-clinical-applications-
and-related-fields-donald-l-schomer-fernando-h-lopes-da-silva/

Functional Analysis Fundamentals and Applications 2nd

Edition Michel Willem

https://ebookmeta.com/product/functional-analysis-fundamentals-
and-applications-2nd-edition-michel-willem/

CLINICAL FUNCTIONAL MRI presurgical functional

neuroimaging 3rd Edition

https://ebookmeta.com/product/clinical-functional-mri-
presurgical-functional-neuroimaging-3rd-edition/

The 21st Century Voice Contemporary and Traditional

Extra Normal Voice 2nd Edition Michael Edward Edgerton

https://ebookmeta.com/product/the-21st-century-voice-
contemporary-and-traditional-extra-normal-voice-2nd-edition-
michael-edward-edgerton/
Fundamentals of Pathology Medical Course and Step 1
Review 2021 Edition Edition Sattar Husain A.

https://ebookmeta.com/product/fundamentals-of-pathology-medical-
course-and-step-1-review-2021-edition-edition-sattar-husain-a/

Bigler's Chronicle of the West: The Conquest of

California, Discovery of Gold, and Mormon Settlement as
Reflected in Henry William Bigler's Diaries Erwin G.
Gudde
https://ebookmeta.com/product/biglers-chronicle-of-the-west-the-
conquest-of-california-discovery-of-gold-and-mormon-settlement-
as-reflected-in-henry-william-biglers-diaries-erwin-g-gudde/

Caleb Jones Blog Writings 1st Edition Dream Hunter

https://ebookmeta.com/product/caleb-jones-blog-writings-1st-
edition-dream-hunter/

Tell Me What Really Happened 1st Edition Chelsea Sedoti

https://ebookmeta.com/product/tell-me-what-really-happened-1st-
edition-chelsea-sedoti/

Beauty and the Bin 1st Edition Joanne O Connell O

Connell Joanne

https://ebookmeta.com/product/beauty-and-the-bin-1st-edition-
joanne-o-connell-o-connell-joanne/
Incubus 2 1st Edition Brandon Varnell

https://ebookmeta.com/product/incubus-2-1st-edition-brandon-
varnell/
 Scott H. Faro
Feroze B. Mohamed
Editors

Functional
Neuroradiology
Principles and Clinical Applications
Second Edition

123
Functional Neuroradiology
Scott H. Faro • Feroze B. Mohamed
Editors

Functional Neuroradiology
Principles and Clinical Applications

Second Edition
Editors
Scott H. Faro Feroze B. Mohamed
Division of Neuroradiology Department of Radiology
Thomas Jefferson University Thomas Jefferson University
Philadelphia, PA, USA Philadelphia, PA, USA

ISBN 978-3-031-10908-9    ISBN 978-3-031-10909-6 (eBook)

https://doi.org/10.1007/978-3-031-10909-6

© Springer Nature Switzerland AG 2012, 2023

This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and
retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter
developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed
to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

A decade has passed since the publication of the first edition of this textbook. During the last
several years, there have been several exciting advances in the field of functional neuroradiol-
ogy and its translation into clinical applications. The relatively new leading neuroradiological
medical society, the American Society of Functional Neuroradiology, has also seen tremen-
dous growth during this period, including a new generation of enthusiastic physicians and MR
physicists working together with the goal of further advancing the role of neuroimaging in
clinical decision making. The field of functional neuroradiology is bright. This second edition
expands upon the core topics covered in the first edition etc.
The new topics include:

CNS Tumor Surveillance and Functional MR Perfusion Imaging

Quantitative and Physiological Magnetic Resonance Imaging in Glioma
Chemical Exchange Saturation Transfer (CEST) Imaging
Functional Neuroradiology of Multiple Sclerosis: Non-BOLD Techniques
Functional Connectivity MR Imaging
Resting State Functional Magnetic Resonance Imaging
Advanced Diffusion Imaging in Neuroradiology
Diffusion Tractography in Neurosurgical Planning: Overview of Advanced Clinical
Applications
Magnetic Resonance Fingerprinting
Radiomics and Radiogenomics in Glioma
Neuroimaging of Brain Tumors in the Era of Radiogenomics
Functional Imaging in Autism Spectrum Disorder
TBI: Sports-Related Injury
Neurological Applications of Magnetic Resonance-Guided Focused Ultrasound Therapy
Anatomical and Functional Features of the Central Nervous System Lymphatic System
Hydrocephalus Imaging
Advanced Neuroimaging for Spine Metastasis
CNS Machine Learning

It is again our hope that this second edition will be a major reference for physicians from
various disciplines, MR physicists, and cognitive neuroscientists in this important growing
field of functional neuroradiology.

Philadelphia, PA, USA Scott H. Faro

Philadelphia, PA, USA  Feroze B. Mohamed

v
Acknowledgments

Our second edition, two volume text, includes updates to the chapters from the first edition as
well as 18 new chapters on advanced functional neuroradiology. I would like to wholeheart-
edly thank all of our 100+ authors who have dedicated all of their precious time to contribute
to our book. These are challenging times for many reasons; however all of these dedicated
colleagues continue to pursue their passion for research and education that will guide and
motivate the next generation of physicians and scientists. Again, I want to acknowledge my
family and all of my colleagues at Jefferson who give me the time and encouragement to pur-
sue my academic interests. This importantly includes my friend and colleague Dr. Feroze
Mohamed as we near our 30 years of collaboration with no end in sight. Last, I would like to
thank Springer Publishing and specifically Maureen Pierce, Margaret Moore, and Arulronika
Pathinathan at Springer for their tremendous help in the preparation of this textbook.

Scott H. Faro

vii
Contents

Part I Diffusion and Perfusion Imaging: Physical Principles

1 
Physical Principles of Diffusion Imaging���������������������������������������������������������������    3
Thinesh Sivapatham and Elias R. Melhem
2 Physical Principles of Dynamic Contrast-­Enhanced
and Dynamic Susceptibility Contrast MRI �����������������������������������������������������������   15
Mark S. Shiroishi, Jerrold L. Boxerman, C. Chad Quarles, Daniel S. R. Stahl,
Saulo Lacerda, Naira Muradyan, Timothy P. L. Roberts, and Meng Law
3 Physical Principles of Non-gadolinium Perfusion Technique
(Arterial Spin Labeling)�������������������������������������������������������������������������������������������   35
Youngkyoo Jung, Huan Tan, and Jonathan H. Burdette

Part II Diffusion and Perfusion Imaging: Clinical Applications

4 Clinical Applications of Diffusion���������������������������������������������������������������������������   49

Juan Márquez, Thiparom Sananmuang, Ashok Srinivasan, Pamela W. Schaefer,
and Reza Forghani
5 
Clinical Applications of MR Perfusion Imaging ��������������������������������������������������� 119
Seyed Ali Nabavizadeh and Ronald L. Wolf
6 
Clinical Application of Perfusion and Diffusion in Stroke ����������������������������������� 161
Tanvir Rizvi and Max Wintermark
7 Clinical Applications of Dynamic Contrast-Enhanced (DCE)
Permeability Imaging����������������������������������������������������������������������������������������������� 175
Saulo Lacerda, Giuseppe Barisano, Mark S. Shiroishi, and Meng Law
8 
CNS Tumor Surveillance and Functional MR Perfusion Imaging����������������������� 201
Amit Desai and Rajan Jain

Part III Magnetic Resonance Spectroscopy and Chemical Exchange

Saturation Transfer Imaging

9 
Magnetic Resonance Spectroscopy: Physical Principles��������������������������������������� 223
Stefan Blüml
10 
Magnetic Resonance Spectroscopy: Clinical Applications����������������������������������� 241
Alena Horská, Adam Berrington, Peter B. Barker, and Ivan Tkáč
11 
Chemical Exchange Saturation Transfer (CEST) Imaging ��������������������������������� 293
Daniel Paech and Lisa Loi

ix
x Contents

Part IV Multimodal Functional Neuroradiology

12 Functional
 Imaging-Based Diagnostic Strategy: Intra-axial
Brain Masses������������������������������������������������������������������������������������������������������������� 311
Arastoo Vossough and Seyed Ali Nabavizadeh
13 Functional
 Neuroimaging of Epilepsy��������������������������������������������������������������������� 345
Noriko Salamon
14 Functional
 Neuroradiology of Traumatic Brain Injury ��������������������������������������� 355
Giacomo Boffa, Eytan Raz, and Matilde Inglese
15 Functional
 Neuroradiology of Multiple Sclerosis:
Non-BOLD Techniques��������������������������������������������������������������������������������������������� 373
Francesca Benedetta Pizzini and Giacomo Talenti
16 Functional
 Neuroradiology of Psychiatric Diseases����������������������������������������������� 393
Paolo Nucifora
17 Neuroimaging of Pain����������������������������������������������������������������������������������������������� 407
Richard H. Gracely and Pia C. Sundgren
18 Quantitative
 and Physiological Magnetic Resonance
Imaging in Glioma ��������������������������������������������������������������������������������������������������� 433
Shah Islam, Melanie A. Morrison, and Adam D. Waldman

Part V BOLD Functional MRI: Physical Principles

19 Principles
 of BOLD Functional MRI ��������������������������������������������������������������������� 461
Seong-Gi Kim and Peter A. Bandettini
20 fMRI Scanning Methods ����������������������������������������������������������������������������������������� 473
Chris J. Conklin and Devon M. Middleton
21 Experimental
 Design and Data Analysis for fMRI ����������������������������������������������� 485
Geoffrey K. Aguirre
22 Challenges
 in fMRI and Its Limitations����������������������������������������������������������������� 497
R. Todd Constable
23 Neurovascular
 Uncoupling in Functional MRI����������������������������������������������������� 511
Jorn Fierstra and David J. Mikulis
24 Functional
 Connectivity MR Imaging ������������������������������������������������������������������� 521
Corey Horien, Xilin Shen, Dustin Scheinost, R. Todd Constable,
and Michelle Hampson
25 Clinical
 Challenges of Functional MRI������������������������������������������������������������������ 543
Nader Pouratian, Bayard Wilson, and Susan Y. Bookheimer

Part VI BOLD Functional MRI: Clinical & Cognitive Applications

26 fMRI
 of Language Systems: Methods and Applications��������������������������������������� 565
Jeffrey R. Binder
27 Functional
 MRI of Language and Memory in Surgical Epilepsy:
fMRI Wada Test ������������������������������������������������������������������������������������������������������� 593
Brenna C. McDonald, Rupa Radhakrishnan, and Kathleen M. Kingery
Contents xi

28 
Resting State Functional Magnetic Resonance Imaging��������������������������������������� 623
Daniel Ryan, Sachin K. Gujar, and Haris I. Sair
29 
fMRI of Human Visual Pathways��������������������������������������������������������������������������� 641
Edgar A. DeYoe, John L. Ulmer, Wade M. Mueller, Lotfi Hacein-Bey,
Viktor Szeder, Mary Jo Maciejewski, Karen Medler, Danielle Reitsma,
and Jedediah Mathis
30 Functional MRI Studies of Memory in Aging, Mild Cognitive
Impairment, and Alzheimer’s Disease ������������������������������������������������������������������� 671
Jian Zhu, Shannon L. Risacher, Heather A. Wishart,
and Andrew J. Saykin
31 
Brain Mapping for Cognitive Neuroscience and Neurosurgery��������������������������� 713
Joy Hirsch
32 
fMRI of the Central Auditory System��������������������������������������������������������������������� 745
Deborah Ann Hall and Thomas M. Talavage
33 fMRI of Epilepsy������������������������������������������������������������������������������������������������������� 765
Karsten Krakow
34 
Functional MRI of Multiple Sclerosis��������������������������������������������������������������������� 781
Heather A. Wishart
35 
Applications of fMRI to Psychiatry������������������������������������������������������������������������� 799
Chandni Sheth, Erin C. McGlade, and Deborah Yurgelun-Todd
36 
Applications of fMRI to Neurodegenerative Disease��������������������������������������������� 819
Shamseldeen Y. Mahmoud, Moon Doksu, Jonathan K. Lee,
and Stephen E. Jones
37 
Applications of MRI to Psychopharmacology������������������������������������������������������� 861
Dan J. Stein, Yihong Yang, and Betty Jo Salmeron
38 
Functional MRI: Cognitive Neuroscience Applications ��������������������������������������� 877
Andrew S. Kayser, Anthony J. W. Chen, and Mark D’Esposito

Part VII Diffusion Tensor Imaging: Physical Principles

39 
Diffusion Tensor Magnetic Resonance Imaging – Physical Principles���������������� 903
Jose Guerrero, Thomas A. Gallagher, Andrew L. Alexander,
and Aaron S. Field
40 
Advanced Diffusion Imaging in Neuroradiology��������������������������������������������������� 933
Devon M. Middleton and Chris J. Conklin

Part VIII Diffusion Tensor Imaging: Clinical Applications

41 
Diffusion Tractography in Neurosurgical Planning: Overview of Advanced Clinical
Applications��������������������������������������������������������������������������������������������������������������� 951
Jingya Miao, Solomon Feuerwerker, Karim Hafazalla, Lauren Janczewski,
Michael P. Baldassari, Steven Lange, Arichena Manmatharayan,
Jennifer Muller, Michael Kogan, Caio M. Matias, Nikolaos Mouchtouris,
Daniel Franco, Joshua E. Heller, James S. Harrop, Ashwini Sharan,
and Mahdi Alizadeh
xii Contents

42 Issues
 in Translating Imaging Technology and Presurgical
Diffusion Tensor Imaging����������������������������������������������������������������������������������������� 969
John L. Ulmer, Jeffrey I. Berman, Wade M. Mueller, Wolfgang Gaggl,
Edgar A. DeYoe, and Andrew P. Klein
43 Clinical
 Applications of Diffusion MRI in Epilepsy����������������������������������������������� 1003
Joanne M. Rispoli, Christopher P. Hess, and Timothy M. Shepherd

Part IX Clinical Presurgical Brain Tumor Mapping

44 fMRI
 and DTI: Review of Complementary Techniques��������������������������������������� 1025
Shruti Agarwal, Hussain Al Khalifah, Domenico Zaca, and Jay J. Pillai
45 DTI:
 Functional Anatomy of Key Tracts ��������������������������������������������������������������� 1061
Arash Kamali, Vinodh A. Kumar, Khader M. Hasan, Mohit Maheshwari,
Andrew P. Klein, Kiran Shankar Talekar, John L. Ulmer, and Scott H. Faro
46 Pediatric Applications of fMRI ������������������������������������������������������������������������������� 1085
Byron Bernal and Nolan R. Altman

Part X Magnetoenphalopathy and PET Imaging

47 Magnetoencephalography:
 Epilepsy and Brain Mapping ����������������������������������� 1123
Erin Simon Schwartz and Timothy P. L. Roberts
48 PET-CT/MR
 Imaging in Head and Neck Cancer: Physiologic
Variations, Pitfalls, and Directed Applications������������������������������������������������������� 1137
Laurie A. Loevner
49 Simultaneous
 PET and MR Imaging of the Human Brain����������������������������������� 1165
Ciprian Catana, Christin Sander, A. Gregory Sorensen, and Bruce R. Rosen

Part XI Emerging Neuroimaging Techniques

50 Functional
 Imaging in Autism Spectrum Disorder����������������������������������������������� 1205
Junko Matsuzaki, Heather Green, and Timothy P. L. Roberts
51 The
 Role of Molecular Imaging in Personalized Medicine����������������������������������� 1223
Michelle Bradbury
52 Quantitative
 Metabolic Magnetic Resonance Imaging:
A Case for Bioscales in Medicine����������������������������������������������������������������������������� 1239
Keith R. Thulborn
53 Magnetic Resonance Fingerprinting����������������������������������������������������������������������� 1259
Chaitra Badve and Dan Ma
54 Neuroimaging
 of Brain Tumors in the Era of Radiogenomics����������������������������� 1275
Prem P. Batchala, Thomas J. Eluvathingal Muttikkal, Joseph H. Donahue,
M. Beatriz Lopes, Eli S. Williams, Nicholas J. Tustison, and Sohil H. Patel
55 Radiomics
 and Radiogenomics in Glioma ������������������������������������������������������������� 1313
Murat Ak and Rivka R. Colen
56 Quantitative T1ρ MR Imaging in Neuroradiology ����������������������������������������������� 1323
Christopher G. Filippi, Alexander Klebba, Scott Hipko, and Richard Watts
Contents xiii

57 Neurological Applications of Magnetic Resonance-Guided

Focused Ultrasound Therapy����������������������������������������������������������������������������������� 1337
Ahmed Abdul-Kareem, Dheeraj Gandhi, Timothy R. Miller, Rao Gullapalli,
and Elias R. Melhem
58 CNS Machine Learning������������������������������������������������������������������������������������������� 1347
Elizabeth Tong, Endre Grøvik, Kyrre Eeg Emblem, Kevin Chen,
Audrey Fan, Yannan Yu, Guangming Zhu, Moss Zhao, Sanaz Niri,
and Greg Zaharchuk
59 Anatomical and Functional Features of the Central Nervous
System Lymphatic System��������������������������������������������������������������������������������������� 1377
Manus Joseph Donahue, Paula M. C. Donahue, Rachelle Crescenzi,
and Colin D. McKnight
60 
TBI Sports Related Injury��������������������������������������������������������������������������������������� 1389
Mohammad I. Kawas, Christopher A. Sheridan, William C. Flood,
Adam P. Sweeney, and Christopher T. Whitlow

Part XII Functional Spine and Hydrocephalus Imaging

61 Functional MRI of the Spinal Cord: Diffusion Weighted, Diffusion

Tensor Imaging, and Fiber Tractography��������������������������������������������������������������� 1403
Kiran Shankar Talekar, Meng Law, Majda M. Thurnher, Eric D. Schwartz,
and Adam E. Flanders
62 
Advanced Neuroimaging for Spine Metastasis������������������������������������������������������ 1425
Varun Sethi, Kristin J. Redmond, and Majid Khan
63 Hydrocephalus Imaging������������������������������������������������������������������������������������������� 1439
Ari M. Blitz, Ameya P. Nayate, Anthony Higginbotham,
Daniele Rigamonti, and Harold L. Rekate

Part XIII Neuroanatomical Brain Atlas

64 Neuroanatomical Atlas of Key Presurgical and Cognitive Eloquent

Cortex Regions ��������������������������������������������������������������������������������������������������������� 1457
Feroze B. Mohamed, Michael P. Yannes, Muhammed Malik,
and Scott H. Faro
65 Normal Anatomic Atlas of Common White Matter
Tracts Using DTI������������������������������������������������������������������������������������������������������� 1477
Andrew P. Klein
66 fMRI Paradigms������������������������������������������������������������������������������������������������������� 1491
Kiran Shankar Talekar, Feroze B. Mohamed, and Scott H. Faro
Index����������������������������������������������������������������������������������������������������������������������������������� 1501
Contributors

Shruti Agarwal, PhD Division of Neuroradiology, Russell H. Morgan Department of

Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore,
MD, USA
Geoffrey K. Aguirre, MD, PhD Department of Neurology, Hospital of the University of
Pennsylvania, Philadelphia, PA, USA
Ahmed Abdul-Kareem, SM, MD Department of Neurosurgery, University of Maryland
School of Medicine, Baltimore, MD, USA
Murat Ak, MD Department of Radiology, UPMC Hillman Cancer Center, Pittsburgh, PA,
USA
Andrew L. Alexander, PhD Departments of Medical Physics and Psychiatry, University of
Wisconsin-Madison, Madison, WI, USA
Mahdi Alizadeh, PhD Department of Neurosurgery, Thomas Jefferson University Hospital,
Philadelphia, PA, USA
Hussain Al Khalifah, MD Division of Neuroradiology, Russell H. Morgan Department of
Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore,
MD, USA
Nolan R. Altman, MD Department of Radiology, Nicklaus Children’s Hospital, Miami, FL,
USA
Chaitra Badve, MBBS, MD Department of Radiology, University Hospitals Cleveland
Medical Center, Case School of Medicine, Cleveland, OH, USA
Michael P. Baldassari, BA Department of Neurological Surgery, Sidney Kimmel Medical
School of Thomas Jefferson University, Philadelphia, PA, USA
Peter A. Bandettini, PhD Section on Functional Imaging Methods, National Institute of
Mental Health, National Institute of Health, Bethesda, MD, USA
Giuseppe Barisano, MD Laboratory of Neuro Imaging, Mark and Mary Stevens
Neuroimaging and Informatics Institute, University of Southern California, Los Angeles, CA,
USA
Peter B. Barker, DPhil Russell H. Morgan Department of Radiology and Radiological
Science, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Prem P. Batchala, MD Department of Radiology and Medical Imaging, University of
Virginia Health System, Charlottesville, VA, USA
Jeffrey I. Berman, PhD Department of Radiology, Children’s Hospital of Philadelphia,
Philadelphia, PA, USA

xv
xvi Contributors

Byron Bernal, MD, PPI Brain Institute—Radiology Department, Nicklaus Children’s Health
System, Miami, FL, USA
Adam Berrington, MSCi, DPhil Sir Peter Mansfield Imaging Centre, School of Physics and
Astronomy, University of Nottingham, Nottingham, UK
Jeffrey R. Binder, MD Neurology and Biophysics, Department of Neurology, Medical
College of Wisconsin, Milwaukee, WI, USA
Ari M. Blitz, MD Division of Neuroradiology, Department of Radiology, University
Hospitals, Case Western Reserve University School of Medicine, Cleveland, OH, USA
Cleveland Medical Center, Cleveland, OH, USA
Stefan Blüml, PhD Department of Radiology and Viterbi School of Engineering, Children’s
Hospital Los Angeles/University of Southern California (USC), Los Angeles, CA, USA
Giacomo Boffa, MD Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics,
Maternal and Child Health (DiNOGMI), University of Genoa, Genoa, Italy
Susan Y. Bookheimer, PhD Department of Psychiatry and Behavioral Sciences, Center for
Autism Research and Treatment, University of California, Los Angeles, Los Angeles, CA,
USA
Jerrold L. Boxerman, MD, PhD, FACR Diagnostic Imaging, Warren Alpert Medical School
of Brown University, Providence, RI, USA
Department of Diagnostic Imaging, Rhode Island Hospital, Providence, RI, USA
Michelle Bradbury, MD, PhD MSK-Cornell Center for Translation of Cancer Nanomedicines,
Intraoperative Imaging Program, Department of Radiology and Molecular Pharmacology
Program, Gerstner Sloan Kettering Graduate School of Biomedical Sciences, Memorial Sloan
Kettering Cancer Center & Weill Medical College of Cornell University, New York, NY, USA
Jonathan H. Burdette, MD Department of Radiology, Wake Forest University School of
Medicine, Winston-Salem, NC, USA
Ciprian Catana, MD, PhD Department of Radiology, Athinoula A. Martinos Center for
Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School,
Charlestown, MA, USA
Anthony J. W. Chen, MD Division of Neurology, Department of Veterans Affairs, Northern
California Health Care System, Martinez, CA, USA
Kevin Chen, PhD Department of Radiology, Stanford University, Stanford, CA, USA
Rivka R. Colen, MD Department of Radiology, UPMC Hillman Cancer Center, Pittsburgh,
PA, USA
Chris J. Conklin, PhD Neuroscience Medical Affairs, Bioclinica, Princeton, NJ, USA
R. Todd Constable, PhD Department of Diagnostic Radiology and Biomedical Imaging,
Biomedical Engineering, and Neurosurgery, Interdepartmental Neuroscience Program, Yale
University School of Medicine, New Haven, CT, USA
MRI Research, Yale MRRC, Yale University School of Medicine, New Haven, CT, USA
Interdepartmental Neuroscience Program, Department of Diagnostic Radiology and
Biomedical Imaging, Biomedical Engineering, and Neurosurgery, Yale University School of
Medicine, New Haven, CT, USA
Rachelle Crescenzi, PhD Department of Radiology and Radiological Sciences, Vanderbilt
University Medical Center, Nashville, TN, USA
Contributors xvii

Mark D’Esposito, MD Division of Neurology, Department of Veterans Affairs, Northern

California Health Care System, Martinez, CA, USA
Helen Wills Neuroscience Institute, University of California at Berkeley, Berkeley, CA, USA
Amit Desai, MD Department of Radiology, Mayo Clinic, Jacksonville, FL, USA
Edgar A. DeYoe, PhD Department of Cell Biology, Neurobiology and Anatomy, Medical
College of Wisconsin, Milwaukee, WI, USA
Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, USA
Department of Radiology, Medical College of Wisconsin, Milwaukee, WI, USA
Moon Doksu, MD Imaging Institute, Cleveland Clinic, Cleveland, OH, USA
Joseph H. Donahue, MD Department of Radiology and Medical Imaging, University of
Virginia Health System, Charlottesville, VA, USA
Manus Joseph Donahue, PhD Department of Radiology and Radiological Sciences,
Vanderbilt University Medical Center, Nashville, TN, USA
Department of Neurology, Vanderbilt University Medical Center, Nashville, TN, USA
Department of Psychiatry, Vanderbilt University Medical Center, Nashville, TN, USA
Paula M. C. Donahue, DPT Department of Physical Medicine and Rehabilitation, Vanderbilt
University Medical Center, Nashville, TN, USA
Thomas J. Eluvathingal Muttikkal, MD Department of Radiology and Medical Imaging,
University of Virginia Health System, Charlottesville, VA, USA
Kyrre Eeg Emblem, PhD Department for Diagnostic Physics, Oslo University Hospital,
Oslo, Norway
Audrey Fan, PhD Department of Biomedical Engineering, University of California Davis,
Davis, CA, USA
Department of Neurology, University of California Davis, Davis, CA, USA
Scott H. Faro, MD, FASFNR Departments of Radiology and Neurology, Division of
Neuroradiology and ENT, Thomas Jefferson University Hospital, Philadelphia, PA, USA
Solomon Feuerwerker, MD Department of Surgery, University of Vermont Medical Center,
Burlington, VT, USA
Aaron S. Field, MD, PhD Department of Radiology, University of Wisconsin School of
Medicine and Public Health, Madison, WI, USA
Jorn Fierstra, MD, PhD Department of Neurosurgery, Clinical Neuroscience Center,
University Hospital Zurich, University of Zurich, Zurich, Switzerland
Christopher G. Filippi, MD Department of Radiology, Tufts Medical Center and Tufts
University School of Medicine, Boston, MA, USA
Adam E. Flanders, MD Department of Radiology/Neuroradiology, Thomas Jefferson
University Hospital, Philadelphia, PA, USA
William C. Flood Department of Neuroscience, Wake Forest School of Medicine, Winston-­
Salem, NC, USA
Reza Forghani, MD, PhD Augmented Intelligence and Precision Health Laboratory
(AIPHL), McGill University Health Center, Montreal, QC, Canada
Combined Research and Clinical Fellowship in Head and Neck Imaging, McGill University
Health Center, Montreal, QC, Canada
xviii Contributors

Daniel Franco, MD Department of Neurosurgery, Thomas Jefferson University Hospital,

Philadelphia, PA, USA
Wolfgang Gaggl, PhD GE Healthcare, Waukesha, WI, USA
Thomas A. Gallagher, MD Department of Radiology, University of Wisconsin School of
Medicine and Public Health, Madison, WI, USA
Dheeraj Gandhi, MD Department of Neurosurgery, University of Maryland School of
Medicine, Baltimore, MD, USA
Department of Diagnostic Radiology & Nuclear Medicine, Neuroradiology, University of
Maryland School of Medicine, Baltimore, MD, USA
Department of Neurology, University of Maryland School of Medicine, Baltimore, MD, USA
Richard H. Gracely, PhD Center for Pain Research and Innovation, Department of
Endodontics, School of Dentistry, School of Medicine, University of North Carolina, Chapel
Hill, Chapel Hill, NC, USA
Heather Green, PhD Department of Radiology, Lurie Family Foundations MEG Imaging
Center, Children’s Hospital of Philadelphia, Philadelphia, PA, USA
Endre Grøvik, PhD Department for Diagnostic Physics, Oslo University Hospital, Oslo,
Norway
Faculty of Health Sciences, University of South-Eastern Norway, Drammen, Norway
Jose Guerrero, PhD Department of Medical Physics, University of Wisconsin-Madison,
Madison, WI, USA
Sachin K. Gujar, MBBS, MD Department of Radiology and Radiological Science, Johns
Hopkins Medicine, Baltimore, MD, USA
Rao Gullapalli, PhD, MBA Department of Diagnostic Radiology & Nuclear Medicine,
Neuroradiology, University of Maryland School of Medicine, Baltimore, MD, USA
Lotfi Hacein-Bey, MD Interventional Radiology, Sutter Health Research Enterprise,
Sacramento, CA, USA
Karim Hafazalla, MD Department of Neurological Surgery, Thomas Jefferson University
Hospital, Philadelphia, PA, USA
Deborah Ann Hall, BSc, PhD Division of Clinical Neuroscience, School of Medicine,
University of Nottingham, Nottingham, UK
Michelle Hampson, PhD Department of Radiology and Biomedical Imaging, Yale University
School of Medicine, New Haven, CT, USA
James S. Harrop, MD Department of Neurosurgery, Thomas Jefferson University Hospital,
Philadelphia, PA, USA
Khader M. Hasan Department of Radiology, University of Texas Houston Medical School,
Houston, TX, USA
Joshua E. Heller, MD, MBA Department of Neurological Surgery, Thomas Jefferson
University Hospital, Philadelphia, PA, USA
Christopher P. Hess, MD, PhD Department of Radiology and Biomedical Imaging,
University of California, San Francisco, San Francisco, CA, USA
Anthony Higginbotham Department of Radiology, University Hospitals, Case Western
Reserve University School of Medicine, Cleveland, OH, USA
Contributors xix

Scott Hipko, BSc Department of Radiology, Robert Larner School of Medicine University of
Vermont, Burlington, VT, USA
Joy Hirsch, PhD Psychiatry, Comparative Medicine, and Neuroscience, Brain Function
Laboratory, Interdepartmental Neuroscience Program, Yale School of Medicine, New Haven,
CT, USA
Haskins Laboratories, New Haven, CT, USA
Neuroscience, Department of Medical Physics and Biomedical Engineering, Faculty of
Engineering Sciences ex officio, University College London, London, UK
Corey Horien, BA, M.Phil Department of Radiology and Biomedical Imaging, Yale
University School of Medicine, New Haven, CT, USA
Alena Horská, PhD Division of Construction and Instruments, National Institutes of Health,
Office of Research Infrastructure Programs, Bethesda, MD, USA
Matilde Inglese, MD, PhD Department of Neuroscience, Rehabilitation, Ophthalmology,
Genetics, Maternal and Child Health (DiNOGMI), University of Genoa, San Martino Polyclinic
Hospital, Genoa, Italy
Shah Islam, MBBS, BSc, FRCR Department of Surgery and Cancer, Imperial College
London, London, UK
Rajan Jain, MD Department of Radiology, New York University Grossman School of
Medicine, New York, NY, USA
Department of Neurosurgery, New York University Grossman School of Medicine, New York,
NY, USA
Lauren Janczewski, MD Department of General Surgery, Northwestern Memorial Hospital,
Chicago, IL, USA
Stephen E. Jones, MD, PhD Division of Neuroradiology, Imaging Institute, Cleveland
Clinic, Cleveland, OH, USA
Youngkyoo Jung, PhD Department of Radiology, University of California, Davis,
Sacramento, CA, USA
Arash Kamali, MD Department of Radiology, University of Texas, Houston, Memorial
Hermann Hospital at Houston Medical Center, Houston, TX, USA
Mohammad I. Kawas, MD Department of Neuroscience, Wake Forest School of Medicine,
Winston-Salem, NC, USA
Andrew S. Kayser, MD, PhD Department of Neurology, University of California at San
Francisco, San Francisco, CA, USA
Division of Neurology, Department of Veterans Affairs, Northern California Health Care
System, Martinez, CA, USA
Helen Wills Neuroscience Institute, University of California at Berkeley, Berkeley, CA, USA
Majid Khan, MD Department of Radiology, Thomas Jefferson University, Philadelphia, PA,
USA
Seong-Gi Kim, PhD Center for Neuroscience Imaging Research/Department of Biomedical
Engineering, Institute for Basic Science/Sungkyunkwan University, Suwon, Korea
Kathleen M. Kingery, PhD Department of Neurology, Indiana University School of
Medicine, Indianapolis, IN, USA
Alexander Klebba, MD Department of Radiology, Donald and Barbara Zucker School of
Medicine at Hofstra-Northwell, Hempstead, NY, USA
xx Contributors

Department of Radiology, Lenox Hill Hospital, New York, NY, USA

Andrew P. Klein, MD Department of Radiology, Neuroradiology, Medical College of
Wisconsin, Froedtert Hospital, Milwaukee, WI, USA
Michael Kogan, MD, PhD Department of Neurosurgery, Thomas Jefferson University
Hospital, Philadelphia, PA, USA
Department of Neurosurgery, University of New Mexico, New Mexico, NM, USA
Karsten Krakow, MD PhD Asklepios Neurologische Klinik Falkenstein, Königstein,
Germany
Vinodh A. Kumar, MD Department of Neuroradiology, The University of Texas MD
Anderson Cancer Center, Houston, TX, USA
Saulo Lacerda, MD Radiologia - Neurorradiologia na Hospital da Bahia, Salvador, Bahia,
Brazil
Steven Lange, MD Department of Radiology, Thomas Jefferson University Hospital,
Philadelphia, PA, USA
Meng Law, MD Nuclear Medicine, Neuroscience, Electrical and Computer Systems
Engineering, Monash University, Alfred Health Organization, Melbourne, VIC, Australia
Department of Radiology and Nuclear Medicine, Monash University, Alfred Hospital,
Melbourne, VIC, Australia
Department of Radiology, Alfred Health, Department of Neuroscience, Monash University,
Melbourne, VIC, Australia
Jonathan K. Lee, MD Imaging Institute, Cleveland Clinic, Cleveland, OH, USA
Laurie A. Loevner, MD Department of Radiology, Division of Neuroradiology, Perelman
Center for Advanced Medicine, Philadelphia, PA, USA
Department of Neurosurgery, Perelman Center for Advanced Medicine, Philadelphia, PA,
USA
Department of Ophthalmology, Perelman Center for Advanced Medicine, Philadelphia, PA,
USA
Penn Radiology Perelman, Perelman Center for Advanced Medicine, Philadelphia, PA, USA
Lisa Loi Department of Radiology, German Cancer Research Center, Heidelberg, Germany
M. Beatriz Lopes, MD, PhD Department of Pathology, Divisions of Neuropathology and
Molecular Diagnostics, University of Virginia Health System, Charlottesville, VA, USA
Mary Jo Maciejewski, PhD Educational Therapy and Enrichment Support, Houston, TX,
USA
Dan Ma, PhD Department of Biomedical Engineering, Case Western Reserve University,
Cleveland, OH, USA
Mohit Maheshwari, MBBS, MD Department of Radiology, Children’s Wisconsin and
Medical College of Wisconsin, Milwaukee, WI, USA
Shamseldeen Y. Mahmoud, MD Department of Radiology, Saint Louis University, St. Louis,
MO, USA
Muhammed Malik, BS Department of Radiology, Temple University School of Medicine,
Philadelphia, PA, USA
Arichena Manmatharayan, MBBS Department of Neuroradiology, Thomas Jefferson
University Hospital, Philadelphia, PA, USA
Contributors xxi

Juan Márquez, MD Diagnostic Imaging, Fundacion Valle Del Lili, Cali, Colombia
Jedediah Mathis, BS Department of Neurology, Medical College of Wisconsin, Milwaukee,
WI, USA
Caio M. Matias, MD Department of Neurosurgery, Thomas Jefferson University Hospital,
Philadelphia, PA, USA
Junko Matsuzaki, PhD Department of Radiology, Lurie Family Foundations MEG Imaging
Center, Children’s Hospital of Philadelphia, Philadelphia, PA, USA
Brenna C. McDonald, PsyD, MBA Department of Radiology and Imaging Sciences, Indiana
University School of Medicine, Indianapolis, IN, USA
Department of Neurology, Indiana University School of Medicine, Indianapolis, IN, USA
Erin C. McGlade, PhD Department of Psychiatry, University of Utah, Salt Lake City, UT,
USA
Colin D. McKnight, MD Department of Radiology and Radiological Sciences, Vanderbilt
University Medical Center, Nashville, TN, USA
Karen Medler Clinical Research Ethics Board, Vancouver Island Health Authority, Victoria,
BC, Canada
Elias R. Melhem, MD, PhD Department of Diagnostic Radiology & Nuclear Medicine,
Neuroradiology, University of Maryland School of Medicine, Baltimore, MD, USA
Jingya Miao, MS Department of Neurosurgery, Thomas Jefferson University, Philadelphia,
PA, USA
Devon M. Middleton, PhD Department of Radiology, Thomas Jefferson University,
Philadelphia, PA, USA
David J. Mikulis, MD Joint Department of Medical Imaging, Toronto Western Hospital,
Toronto, ON, Canada
Timothy R. Miller, MD Department of Diagnostic Radiology & Nuclear Medicine,
Neuroradiology, University of Maryland School of Medicine, Baltimore, MD, USA
Feroze B. Mohamed, PhD, FASFNR Department of Radiology, Thomas Jefferson University,
Philadelphia, PA, USA
Melanie A. Morrison, PhD Department of Radiology and Biomedical Imaging, University
of California San Francisco, San Francisco, CA, USA
Department of Medicine, Imperial College London, London, UK
Nikolaos Mouchtouris, MD Department of Neurosurgery, Thomas Jefferson University
Hospital, Philadelphia, PA, USA
Wade M. Mueller, MD Department of Neurosurgery, Medical College of Wisconsin,
Milwaukee, WI, USA
Jennifer Muller, MSc Department of Neurosurgery, Thomas Jefferson University Hospital,
Philadelphia, PA, USA
Naira Muradyan, PhD U.S. Food and Drug Administration, Center for Devices and
Radiological Health, Silver Spring, MD, USA
Seyed Ali Nabavizadeh, MD Department of Radiology, Division of Neuroradiology,
Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA
Ameya P. Nayate, MD Cleveland Medical Center, Cleveland, OH, USA
xxii Contributors

Department of Radiology, University Hospitals, Case Western Reserve University School of

Medicine, Cleveland, OH, USA
Sanaz Niri, MD Department of Radiology, Stanford University, Stanford, CA, USA
Paolo Nucifora, MD, PhD Department of Radiology, Loyola University Chicago, Maywood,
IL, USA
Daniel Paech, MD, MS Department of Radiology, German Cancer Research Center,
Heidelberg, Germany
Sohil H. Patel, MD Department of Radiology and Medical Imaging, University of Virginia
Health System, Charlottesville, VA, USA
Jay J. Pillai, MD Division of Neuroradiology, Mayo Clinic College of Medicine and Science,
Rochester, MN, USA
Francesca Benedetta Pizzini, MD, PhD Radiology, Department of Diagnostic and Public
Health, Verona University, Verona, Italy
Nader Pouratian, MD, PhD Department of Neurological Surgery, University of California,
Los Angeles, Los Angeles, CA, USA
C. Chad Quarles, PhD Division of Neuroimaging Research, Barrow Neurological Institute,
St. Joseph’s Hospital and Medical Center, Phoenix, AZ, USA
Rupa Radhakrishnan, MBBS, MS Department of Radiology and Imaging Sciences,
Indiana University School of Medicine, Indianapolis, IN, USA
Eytan Raz, MD, PhD Department of Radiology, New York University Medical Center, New
York, NY, USA
Kristin J. Redmond, MD, MPH Department of Radiation Oncology and Molecular Radiation
Sciences, The Johns Hopkins University, Baltimore, MD, USA
Danielle Reitsma, PhD Department of Radiology, Medical College of Wisconsin, Milwaukee,
WI, USA
Harold L. Rekate Department of Neurosurgery, Hofstra Northwell School of Medicine,
Great Neck, NY, USA
Department of Neurosurgery, Northshore University Hospital, Manhasset, NY, USA
Daniele Rigamonti Department of Neurosurgery, Johns Hopkins Medical Institutions,
Baltimore, MD, USA
Shannon L. Risacher, PhD Department of Radiology and Imaging Sciences, Indiana
Alzheimer’s Disease Research Center and Center for Neuroimaging, Indiana University
School of Medicine, Indianapolis, IN, USA
Joanne M. Rispoli, MD Department of Radiology, New York University, New York, NY,
USA
Department of Radiology, Boston Children’s Hospital, Boston, MA, USA
Tanvir Rizvi, MBBS, MD Department of Radiology and Medical Imaging, University of
Virginia Health, Charlottesville, VA, USA
Timothy P. L. Roberts, PhD Department of Radiology, Lurie Family Foundations MEG
Imaging Center, Children’s Hospital of Philadelphia, Philadelphia, PA, USA
Department of Radiology, Children’s Hospital of Philadelphia, Philadelphia, PA, USA
Bruce R. Rosen, MD, PhD Department of Radiology, Athinoula A. Martinos Center for
Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School,
Charlestown, MA, USA
Contributors xxiii

Daniel Ryan, MD Department of Neuroradiology, Johns Hopkins Hospital, Baltimore, MD,

USA
Haris I. Sair, MD Department of Radiology and Radiological Science, Johns Hopkins
Medicine, Baltimore, MD, USA
The Malone Center for Engineering in Healthcare, The Whiting School of Engineering, Johns
Hopkins University, Baltimore, MD, USA
Noriko Salamon, MD, PhD Department of Radiology, David Geffen School of Medicine at
UCLA, Ronald Reagan Medical Center, Los Angeles, CA, USA
Betty Jo Salmeron, MD Cognitive and Affective Neuroscience of Addiction Section,
Biomedical Research Center, The National Institute on Drug Abuse (NIDA), Baltimore, MD,
USA
Thiparom Sananmuang, MD Department of Diagnostic and Therapeutic Radiology and
Research, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
Christin Sander, PhD Department of Radiology, Athinoula A. Martinos Center for
Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School,
Charlestown, MA, USA
Andrew J. Saykin, PsyD Department of Radiology and Imaging Sciences, Indiana
Alzheimer’s Disease Research Center and Center for Neuroimaging, Indiana University
School of Medicine, Indianapolis, IN, USA
Pamela W. Schaefer, MD Harvard Medical School, Massachusetts General Hospital, Boston,
MA, USA
Dustin Scheinost, PhD Interdepartmental Neuroscience Program, Department of Radiology
and Biomedical Imaging, Department of Biomedical Engineering, Department of Statistics &
Data Science, Child Study Center, Yale University School of Medicine, New Haven, CT, USA
Eric D. Schwartz, MD Department of Radiology, St. Elizabeth’s Medical Center, Brighton,
MA, USA
Erin Simon Schwartz, MD, FACR Department of Radiology, Perelman School of Medicine,
University of Pennsylvania, Philadelphia, PA, USA
Varun Sethi, MD Department of Radiology, Temple University, Philadelphia, PA, USA
Ashwini Sharan, MD Department of Neurosurgery, Thomas Jefferson University Hospital,
Philadelphia, PA, USA
Xilin Shen, PhD Department of Radiology and Biomedical Imaging, Yale University School
of Medicine, New Haven, CT, USA
Timothy M. Shepherd, MD, PhD Department of Radiology, New York University, New
York, NY, USA
Christopher A. Sheridan Department of Neuroscience, Wake Forest School of Medicine,
Winston-Salem, NC, USA
Chandni Sheth, PhD Department of Psychiatry, University of Utah, Salt Lake City, UT, USA
Mark S. Shiroishi, MD, MS Division of Neuroradiology, Department of Radiology, Keck
School of Medicine of University of Southern California, Los Angeles, CA, USA
Thinesh Sivapatham, MD Comprehensive Stroke Program, Interventional Neuroradiology,
Christiana Care Health System, Newark, DE, USA
xxiv Contributors

A. Gregory Sorensen, MD Department of Radiology, Athinoula A. Martinos Center for

Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School,
Charlestown, MA, USA
Ashok Srinivasan, MD, FACR University of Michigan, Ann Arbor, MI, USA
Daniel S. R. Stahl, DO Rutgers New Jersey Medical School, Newark, NJ, USA
Dan J. Stein, PhD Department of Psychiatry and Mental Health, University of Cape Town,
Cape Town, South Africa
Pia C. Sundgren, MD, PhD Department of Diagnostic Radiology, Clinical Sciences Lund,
Lund University, Lund, Sweden
Department for Medical Imaging and Physiology, Skåne University Hospital, Lund, Sweden
Adam P. Sweeney Department of Radiology, Wake Forest School of Medicine, Winston-­
Salem, NC, USA
Viktor Szeder, MD, PhD, MSc Radiology and Neurosurgery, Division of Interventional
Neuroradiology, Department of Radiology, UCLA Ronald Reagan Medical Center, Los
Angeles, CA, USA
Thomas M. Talavage, PhD Department of Biomedical Engineering, College of Engineering
and Applied Science, University of Cincinnati, Cincinnati, OH, USA
Kiran Shankar Talekar, MBBS, MD, DABR Department of Radiology, Division of
Neuroradiology and ENT, Thomas Jefferson University Hospital, Philadelphia, PA, USA
Giacomo Talenti, MD Neuroradiology Unit, University Hospital of Padua, Padua, Italy
Huan Tan, PhD Beghou Consulting, Durham, NC, USA
Keith R. Thulborn, MD, PhD Center for Magnetic Resonance Research, University of
Illinois Health, Chicago, IL, USA
Majda M. Thurnher, MD Department of Biomedical Imaging and Image-Guided Therapy,
Medical University of Vienna, Vienna, Austria
Ivan Tkáč, PhD Center for Magnetic Resonance Research, University of Minnesota,
Minneapolis, MN, USA
Elizabeth Tong, MD Department of Radiology, Stanford University, Stanford, CA, USA
Nicholas J. Tustison, DSc Department of Radiology and Medical Imaging, University of
Virginia Health System, Charlottesville, VA, USA
John L. Ulmer, MD Department of Radiology, Medical College of Wisconsin, Milwaukee,
WI, USA
Neuroradiology, Medical College of Wisconsin, Milwaukee, WI, USA
Arastoo Vossough, PhD, MD Department of Radiology, Children’s Hospital of Philadelphia—
University of Pennsylvania, Philadelphia, PA, USA
Adam D. Waldman, BSc (Hons), MA, MB, BChir, PhD Centre for Clinical Brain Sciences,
University of Edinburgh, Edinburgh, UK
Richard Watts, DPhil Department of Psychology, Yale University School of Medicine, New
Haven, CT, USA
Christopher T. Whitlow, MD, PhD, MHA Department of Radiology, Wake Forest School of
Medicine, Winston-Salem, NC, USA
Contributors xxv

Eli S. Williams, PhD Department of Pathology, Division of Laboratory Medicine, University

of Virginia Health System, Charlottesville, VA, USA
Bayard Wilson, MD Department of Neurological Surgery, University of California, Los
Angeles, Los Angeles, CA, USA
Max Wintermark, MD, MS, MBA Department of Neuroradiology, The University of Texas
MD Anderson Cancer Center, Houston, TX, USA
Heather A. Wishart, PhD Department of Psychiatry, Geisel School of Medicine at
Dartmouth, Lebanon, NH, USA
Department of Psychiatry, Dartmouth-Hitchcock Medical Center, Geisel School of Medicine
at Dartmouth, Lebanon, NH, USA
Ronald L. Wolf, MD, PhD Department of Radiology, Division of Neuroradiology, Perelman
School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA
Yihong Yang, PhD Neuroimaging Research Branch, Magnetic Resonance Imaging and
Spectroscopy Section, Biomedical Research Center, The National Institute on Drug Abuse
(NIDA), Baltimore, MD, USA
Michael P. Yannes, MD Diagnostic Radiology, St. Luke’s University Health Network, The
Vascular Center, Easton, PA, USA
Deborah Yurgelun-Todd, PhD Department of Psychiatry, University of Utah, Salt Lake
City, UT, USA
Yannan Yu, MD Department of Radiology, Stanford University, Stanford, CA, USA
Domenico Zaca, PhD Siemens Healthcare, Milan, Italy
Greg Zaharchuk, MD, PhD Department of Radiology, Stanford University, Stanford, CA,
USA
Moss Zhao, DPhil Department of Radiology, Stanford University, Stanford, CA, USA
Guangming Zhu, MD, PhD Department of Radiology, Stanford University, Stanford, CA,
USA
Jian Zhu, PhD Department of Radiology and Imaging Sciences, Indiana Alzheimer’s Disease
Research Center and Center for Neuroimaging, Indiana University School of Medicine,
Indianapolis, IN, USA
Department of Psychology, Eastern Illinois University, Charleston, IL, USA
 Part I
Diffusion and Perfusion Imaging:
Physical Principles
 Physical Principles of Diffusion Imaging
1
Thinesh Sivapatham and Elias R. Melhem

Introduction sexual cells of plants, but after further investigation of both

organic and inorganic materials, he found this motion to be a
While a still body of water may appear static to the eye, general property of small particles suspended in solution. We
water molecules are in constant random motion at a micro- currently know that this motion is attributed to the constant
scopic level. This is termed Brownian motion, and is a result motion of the water molecules that the particles are sus-
of the thermal agitation of the water molecules. While pended in. This motion is known as Brownian motion, or
Brownian motion is a microscopic phenomenon, it results in diffusion, and is a result of constant random microscopic
a macroscopically observable phenomenon known as diffu- molecular motion due to thermal agitation. The molecular
sion. The diffusion of water molecules in the brain provides motion is related to the thermal kinetic energy (Ekin), of the
us with a sensitive window to its underlying physiology and molecules, which is proportional to the temperature, T: Ekin
structure. Diffusion-weighted imaging (DWI) of the brain = (3/2)kBT; where kB = 1.38 × 10−23 J/K is the Boltzmann
was introduced into clinical use in the early 1990s, primarily constant. This Brownian motion also depends upon the size
in the detection of acute ischemic stroke [1–4]. Since that of the particles, density, and viscosity of the medium.
time, advances in technology have resulted in significant Diffusion is a naturally occurring transport process at the
improvements in image quality, allowing the application of molecular level that describes the spread of particles through
DWI to the evaluation of a variety of intracranial disease pro- random motion. This mixing or mass transport occurs as a
cesses, such as infections, neoplasms, demyelinating pro- result of collisions between molecules in liquids and gases
cesses, and trauma. The development of diffusion tensor rather than by bulk motion as is necessary for other transport
imaging (DTI) has allowed mapping of white matter tracts in mechanisms such as convection or dispersion. When there is
the brain, and is discussed elsewhere in this book. In this a concentration gradient, particles spread from areas of
chapter, we review the physical principles of DWI, while the higher concentration to areas of lower concentration until
clinical applications of DWI are discussed in following their distribution becomes equilibrate. This results in a net
chapters. flux of particles with a magnitude that is proportional to the
concentration gradient and to the diffusion coefficient, as
described by Fick’s First Law and represented by the follow-
Brownian Motion and Molecular Diffusion ing equation:
J = D∗∇ C (1.1)
In 1827, Robert Brown, a Scottish botanist, discovered the
random and constant motion of pollen grains suspended in where J is the flux density, D is diffusion coefficient, and ∇C
water while studying them under a light microscope [5]. is change in concentration. The diffusion coefficient is a con-
Brown initially believed this motion to be related to the male stant for a given substance in a given medium of known vis-
cosity at a given temperature. Diffusion can also occur in the
absence of a concentration gradient, but in this scenario the
T. Sivapatham diffusive fluxes cancel each other out, resulting in no net
Comprehensive Stroke Program, Interventional Neuroradiology,
flux.
Christiana Care Health System, Newark, DE, USA
The concept of molecular diffusion was first formally
E. R. Melhem (*)
described by Einstein in 1905 [6]. In a contained volume of
Department of Diagnostic Radiology and Nuclear Medicine,
University of Maryland School of Medicine, Baltimore, MD, USA water, each water molecule undergoes random motion as
e-mail: [email protected] part of the diffusion process. This phenomenon of thermal

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 3

S. H. Faro, F. B. Mohamed (eds.), Functional Neuroradiology, https://doi.org/10.1007/978-3-031-10909-6_1
4 T. Sivapatham and E. R. Melhem

motion of water molecules in a medium that itself consists water, and only the “apparent diffusion coefficient” (ADC)
primarily of water is termed self-diffusion. The path of any can be calculated [7–9]. The average ADC in the brain is
single water molecule is completely random and would be approximately 0.7 × 10-3 mm2/s [10], about four times
impossible to predict, limited only by the boundaries of the smaller than the diffusion coefficient in free water.
container. Einstein was able to prove that the squared dis-
placement of molecules from their starting point could be
described by the equation:  iffusion Encoding and the Stejskal-Tanner
D
Equation
r 2 = 6 Dt (1.2)
where <r2> refers to the mean squared displacement of the Shortly after Bloch and Purcell discovered the NMR phe-
molecules from their original location, t is the diffusion time, nomenon [11–13], Hahn reported his findings that the NMR
and D is the diffusion coefficient for the particular substance spin-echo was sensitive to the effects of diffusion [14]. He
being studied. The diffusion coefficient is typically expressed noted that the random thermal motion of the spins would
in units of mm2/s and relates the average displacement of a reduce the amplitude of the observed spin-echo signal as a
molecule over an area to the observation time; higher diffu- result of the dephasing that occurs in the presence of mag-
sion constants infer increased mobility of water molecules. netic field inhomogeneity, which results in local magnetic
The diffusion coefficient of water at body temperature (37 field gradients. Building on these observations, Carr and
°C) is 3 × 10−3 mm2/s. The distribution of squared displace- Purcell proposed NMR sequences using constant gradients
ments of free water molecules is typically a Gaussian (bell-­ to sensitize the NMR spin-echo to the effects of diffusion,
shaped) function with peak at zero displacement, indicating and developed a mathematical framework to measure the dif-
that most molecules have the same position at the starting fusion coefficient from these sequences [15]. In 1956, Torrey
point and at time t [7, 8]. The probability of molecular dis- modified Bloch’s magnetization equations to include the
placement by a given distance from the starting position is effects of molecular diffusion [16]; the Bloch-Torrey equa-
the same regardless of the direction of measurement, with tions describe how net magnetization depends on several fac-
standard deviation proportional to the diffusion coefficient tors, including longitudinal and transverse magnetization, as
and time measured. well as diffusion.
Diffusion-weighted (DW) magnetic resonance imaging In their seminal paper on the spin-diffusion experiment in
(MRI) utilizes Brownian motion to study the movement of 1965, Stejskal and Tanner described the theory of the pulsed
water in vivo, thereby garnering information about the gradient spin-echo (PGSE), which replaced the steady-state
underlying tissue structure of the brain. DW MRI does not gradients used by Carr and Purcell with short-duration gradi-
measure the diffusion coefficient directly, but rather the ent pulses [17]. This resulted in much improved sensitivity to
effect of the mean displacement of water molecules within diffusion by distinguishing the encoding time (pulse dura-
each three-dimensional (3-D) volume element, or voxel, on tion) and the diffusion time. This also allowed the direct
the nuclear magnetic resonance (NMR) signal. One might measurement of the diffusion function and paved the way for
ask: If the diffusion coefficient of water at body temperature quantitative measurements of the diffusion coefficient. In the
is a constant, then how can we use this information to evalu- presence of a magnetic field, static spins accumulate phase
ate tissue structure? We must remember that in vivo, water shifts according to the equation:
molecules do not freely diffuse as they would in a medium of
Φ ( t ) = γ B0t + ∫ G ( t ) . X ( t ) dt (1.3)
water alone. In tissues, the movement of water occurs largely
in the extracellular space, and their movement is modified by The first term (γB0t) in the equation represents the phase
interactions with cell membranes and macromolecules, i.e., accumulation owing to the static magnetic field, while the
the underlying tissue structure of the brain. Additionally, the second term (∫G(t).X(t)dt) reflects the effect of a magnetic
movement of water molecules in vivo due to diffusion cannot field gradient. The phase accumulation owing to the gradient
be distinguished from the movement of water molecules is proportional to the strength of the field gradient, spatial
from other sources such as active transport, pressure gradi- location of the spin, and the duration of the gradient pulse.
ents, ionic interactions, and changes in membrane permea- In their experiment, Stejskal and Tanner sensitized a
bility. As a result, the overall movement of water molecules standard spin-echo T2-weighted pulse sequence to diffu-
in the brain is reduced as compared to their movement in free sion by adding a pair of diffusion gradients along the same
1 Physical Principles of Diffusion Imaging 5

axis both before and after the 180° refocusing pulse [17]. The effects of signal loss due to diffusion can be explained
The basic idea is to excite the spins with a 90° radiofre- by the Stejskal-Tanner equation:
quency (RF) pulse, encode the spin position with a time- S = S0 exp ( −bD ) (1.4)
constant magnetic field gradient of duration δ, invert the
spin phase with a 180° refocusing pulse, apply a second where S is the signal intensity observed in a given voxel with
magnetic field gradient equal in intensity and duration to a diffusion gradient applied, S0 is the signal intensity of the
the first gradient at a time Δ after the first gradient, and then same voxel in the absence of a diffusion gradient, b is the
acquire the echo at time TE—echo time. The application of diffusion sensitivity factor, and D is the diffusion coefficient.
a linear gradient to a homogeneous magnetic field results in The diffusion sensitivity factor (b) is a function of the
phase shifts of the spins along that axis, which is position strength, duration, and temporal spacing of the diffusion-­
dependent; this phase dispersion leads to signal loss. sensitizing gradients, and can be expressed by the following
However, the application of a second gradient equal to the equation:
first in magnitude and duration but opposite in direction can b = γ 2 G 2δ 2 ( ∆ − δ / 3 ) (1.5)
refocus the phase changes. The first gradient is called the
dephasing gradient, while the second is called the rephas- In this equation, γ is the gyromagnetic ratio (the ratio of
ing gradient [18]. magnetic moment to angular moment of a nuclear spin, a
If a spin is stationary between the applications of the two constant), G is the amplitude of the diffusion gradient (usu-
gradient pulses, the net effect of the gradients is zero and the ally measured in milliteslas per meter), δ is the duration of
spin maintains its signal intensity. This explains the rela- the diffusion gradient pulse (measured in milliseconds), and
tively high signal intensity on DWI seen in the setting of Δ the time interval between the dephasing and rephasing gra-
cytotoxic edema, where there is a relative increase in intra- dient pulses (also measured in milliseconds). The diffusion
cellular water content and water molecules are “trapped” sensitivity factor (b) is measured in units of seconds/mm2.
inside the cells (i.e., unable to diffuse freely) and relative Raising the b-value increases the degree of diffusion weight-
decrease in the size of the extracellular compartment [19– ing (i.e., increases the signal loss caused by the diffusion of
25]. For a spin that moves along the axis of the diffusion water molecules along the direction that the gradient was
gradient, phase accumulation due to the two gradients is no applied; Fig. 1.1). The most effective ways to increase diffu-
longer equal so the rephasing is incomplete. As a result, sion sensitivity, as seen in Eq. (1.3), are to increase gradient
intravoxel dephasing occurs and there is a loss of signal amplitude (G) and gradient duration (δ), as these parameters
intensity; the greater the distance the spin moves along the have a quadratic effect on b. Typical b-values in clinical DWI
axis of the gradient, the greater the signal loss [8, 14, 16, 17]. range from 0 to 1500 s/mm2. However, it should be empha-
This explains the relatively low signal intensity seen on DWI sized that higher diffusion weightings increase exponentially
in the normal brain parenchyma, and even lower signal inten- the contrast between tissues with different diffusion coeffi-
sity in the cerebrospinal fluid (CSF) spaces, where water can cients, but at the same time they decrease the overall signal
diffuse the most freely. intensity and signal-to-noise ratio (SNR) [26, 27].
6 T. Sivapatham and E. R. Melhem

Fig. 1.1 Diffusion-weighted

images acquired using
diffusion sensitivity factors
(b) of increasing strength. The
b = 0 images (top row)
demonstrate the T2-weighting
inherent in DWI. As the
diffusion weighting is
increased (b = 500 and b =
1000, second and third rows,
respectively), the diffusion of
water molecules results in
signal loss, most apparent in
the CSF where water can
diffuse freely. Corresponding
ADC map is shown in the
bottom row. ADC apparent
diffusion coefficient, CSF
cerebrospinal fluid, DWI
diffusion-weighted imaging

Diffusion Anisotropy Biological tissues are heterogeneous in their structure and

consist of various compartments and barriers of various per-
Isotropic diffusion refers to a condition where molecular meability. Neuronal tissue, in particular, is highly fibrillar,
motion is the same in all directions. Since the diffusivity is with tightly packed and coherently aligned axons surrounded
independent of direction in this scenario, the displacement by a network of glial cells that are often themselves orga-
distribution is Gaussian and can be conceptualized as a nized in bundles. The movement of water molecules in these
sphere (i.e., if the center of the sphere is the starting point of tissues is not isotropic, but rather has a propensity to be
a water molecule, the probability of diffusing any given dis- greater in a direction parallel to the fiber tracts than perpen-
tance from the center is equal in all directions) [7, 8, 28]. dicular to it. Diffusion in this scenario varies with direction,
1 Physical Principles of Diffusion Imaging 7

and is described as anisotropic [29–33]. The associated dis- b-value and high b-value (commonly 0 s/mm2 and 1000 s/
placement distribution of anisotropic diffusion is not mm2 in clinical practice). The low b-value sequence essen-
Gaussian or spherical, but rather ellipsoid (or even more tially yields a T2-weighted image. The DW image Si (with b
complex if the underlying tissues contain fibers with various = 1000 s/mm2) can then be divided by the T2-weighted
orientations). This anisotropy can be explained by the diffu- image S0 (with b = 0 s/mm2) to remove the T2-weighting
sion tensor model, and is discussed further in the chapter on effects and produce an exponential image, also known as an
diffusion tensor imaging (DTI). attenuation coefficient map, exponential diffusion-weighted
While isotropic diffusion can be seen in the cerebrospinal image, or the attenuation factor map (shown in Fig. 1.3).
fluid (CSF) spaces and in the gray matter of the adult cere- Areas of restricted diffusion on these maps will demonstrate
bral cortex, diffusion in the white matter is primarily aniso- increased signal intensity, similar to DWI.
tropic. Multiple explanations have been suggested for the ADC values can also be calculated by solving for ADC in
mechanisms underlying diffusion anisotropy of white mat- Eq. (1.7):
ter, including the myelin sheath, axonal direction, axonal ADCi = − ln ( Si / S0 ) / b (1.7)
transport, and local susceptibility gradients [34–36]. Myelin
itself does not appear to be necessary for diffusion anisot- where ln is the natural logarithm. Instead of deriving the
ropy, as experiments have shown anisotropic diffusion in the ADC value mathematically, however, the ADC is typically
absence of myelin [37, 38]. A series of experiments in the determined graphically by obtaining two image sets (again,
1990s excluded the effects of susceptibility-induced gradi- low b-value and high b-value image sets) and plotting the
ents, axonal cytoskeleton, and fast-axonal transport as the natural logarithm of Si versus b for the two b-values; the
etiology of anisotropy in white matter [39–41]. Current evi- ADC is then determined from the slope of that line.
dence suggests that the presence of intact cell membranes is Biological tissue can be considered a combination of
the tissue component predominantly responsible for the intra- and extracellular compartments in which the water
anisotropy of molecular diffusion in white matter, while the molecule is in a state of continuous exchange between these
degree of myelination and cellular density serve to modulate two compartments. The observed signal attenuation in the
anisotropy [42]. diffusion experiment therefore depends on the rate of
exchange and the diffusion time. In the limit of slow
exchange, water spins remain within their respective com-
partments during the diffusion time and signal attenuation
Calculating the ADC
follows a bi-exponential behavior. This bi-exponential
behavior is given by the equation:
Since diffusion in the brain is approximated by the ADC
rather than by direct measurement of the diffusion coeffi- S = S0  f1 exp ( −b1ADC1 ) + f 2 exp ( −b2 ADC2 )  (1.8)
cient, Eq. (1.2) might be better represented as:
where f1 and f2 are the volume fractions of water spins within
Si = S0 exp ( −bADCi ) (1.6)
each of the two compartments and f1 + f2 = 1; ADC1 and
where Si is the signal intensity in a given voxel with the ADC2 are the apparent diffusion coefficients in the two
diffusion-­sensitization gradient applied along direction i, and compartments.
ADCi is the ADC in the i direction. In an isotropic environ- On the other hand, in the limit of fast exchange with com-
ment, the direction of the gradient pulse is irrelevant, since plete redistribution of water between the two compartments
the ADCi would be identical for all directions i, and a single during the diffusion time, the signal attenuation follows a
(scalar) diffusion gradient application is sufficient to calcu- single exponential behavior given by the equation:
late the ADC. Higher ADC values result in lower signal S / S0 = exp ( −b∗ ADC ) (1.9)
intensity Si in the DWI, while reduced ADC values result in
higher signal intensity. In addition to contrast due to differ- The observed apparent diffusion coefficient for a two-­
ences in ADC, the long TE of a DWI pulse sequence makes compartment system includes contributions from both the
it inherently sensitive to T2-contrast as well. T2-prolongation intracellular and extracellular environments and is approxi-
in pathologic tissues can elevate the DWI signal intensity in mated by the equation:
the absence of reduced ADC values [43]. This “T2-shine-­
ADC = f1ADC1 + f 2 ADC2 (1.10)
through” effect results in the DWI hyperintensity being less
specific than reduced ADC values on ADC maps in the mea- In most practical situations, the bi-exponential behavior
surement of true restricted diffusion in tissues (Fig. 1.2). of signal attenuation is not observed. Therefore, it is a com-
In order to remove the T2-effects, the diffusion experi- mon practice to fit the DWI data to the single exponential
ment is repeated for each gradient direction with a low decay model.
8 T. Sivapatham and E. R. Melhem

a b

c d

Fig. 1.2 Images from a patient who presented with an acute stroke restricted diffusion, but rather T2-shine-through on DWI. The lesion
demonstrate the T2-shine-through effects of DWI. The FLAIR image marked by the arrowhead in (d), on the contrary, also demonstrates
(d) shows multifocal areas of T2-prolongation in the periventricular and marked signal intensity on the b = 0 and b = 1000 images, but is dark on
deep white matter (arrow and arrowhead). The lesion marked by the the ADC map, indicating that this is an area of restricted diffusion. ADC
arrow in (d) also demonstrates increased signal on the b = 0 (a), b = apparent diffusion coefficient, FLAIR fluid-attenuated inversion recov-
1000 (b), and ADC (c) images, indicating that this is not an area of ery, DWI diffusion-weighted imaging
1 Physical Principles of Diffusion Imaging 9

Fig. 1.3 Diffusion

anisotropy. The top three rows
show apparent diffusion
coefficient (ADC) maps
derived from diffusion
gradients applied in three
orthogonal directions.
Variations in brightness
between the three sets of
images demonstrate the
anisotropic nature of diffusion
in the brain; diffusion of
water molecules is greater
along nerve bundles than
perpendicular to them. When
the gradient is applied in the x
direction (Dxx; diffusion
gradient in the left-to-right
direction), diffusion is higher
(brighter signal) in fibers that
course in the left-to-right
direction (i.e., splenium) than
perpendicular to that direction
(i.e., posterior limb internal
capsule, and anterior-to-­
posterior-oriented white
matter tracts of the cerebral
hemispheres). With gradient
applied in the y direction
(Dyy; anterior to posterior),
diffusion is greatest in the
anterior-to-posterior-oriented
cerebral white matter tracts,
and lowest in the
perpendicularly oriented
splenium and posterior limb
internal capsule. With
gradient applied in the z
direction (Dzz; cranial to
caudal), diffusion is greatest
in the posterior limb internal
capsule oriented in the same
direction, and lowest in the
perpendicularly oriented
splenium and anterior-to-­
posterior cerebral white
matter tracts. The trace ADC
images shown in the fourth
row are the average of the
ADCs from each of the three
directions, and show diffusion
magnitude information alone
with directional information
removed. The fifth row shows
the corresponding isotropic
diffusion-weighted (DW)
images. The exponential DW
images (Exp, bottom row)
were derived by dividing the
b = 1000 images by the b = 0
images (with b measured in
seconds/mm2) in order to
remove the T2-weighting
effects
10 T. Sivapatham and E. R. Melhem

Isotropic DWI and Trace ADC tant to concomitantly view the ADC map or exponential
image to evaluate for true restricted diffusion.
Diffusion gradient pulses are applied in one direction at a In neonates and young children, the ADC is initially
time, with the resulting DW image giving both directional much higher in rapidly developing gray matter structures
and magnitude information about the ADC. This would be like the thalamus and basal ganglia, and can be more than
sufficient if imaging an isotropic tissue, since the ADC twice as high in the slowly developing white matter regions
would be the same regardless of the direction of the diffusion [44, 45]. This can be attributed to the relatively high water
gradient. When imaging an anisotropic tissue like the white content in the immature brain, with relative paucity of
matter of the brain, however, interpretation of the DWI myelinated neurons. ADC values decrease rapidly over the
would be challenging if only a single gradient direction were first 2 years of brain development, and continue to decrease
probed, due to the variable signal intensity of white matter gradually through childhood, adolescence, and into young
tracts depending on their orientation relative to the direction adulthood [46–52]. Due to the higher ADC values in neo-
of the diffusion gradient. While the diffusion anisotropy dis- nates and infants compared to adults [45–48], it is common
cussed in the previous section can be exploited to image fiber to reduce the diffusion sensitivity factor (b) in this age
tracts in DTI, the directional effects of anisotropy are unde- group to 600 or 700 s/mm , compared to the 1000 s/mm
2 2

sirable in routine clinical DWI. typically used in adults. In the aging but otherwise healthy
In order to overcome the directional influences of the dif- brain, mild increases can be seen in the ADC, particularly
fusion gradient in anisotropic tissues, four separate scalar in white matter [53–58]. Engelter et al. found a significant
image acquisitions are required: one without a diffusion-­ correlation between the average ADC of white matter and
sensitizing gradient (S0, where b = 0 s/mm2), and three with age, with patients 60 years of age or older having increased
the diffusion gradients applied in three orthogonal directions ADC compared to those under the age of 60; a similar trend
x, y, and z (which will be called Sx, Sy, and Sz). To create an was seen in the average ADC of the thalamus [55]. Chen
image related only to the magnitude of the ADC, the DW et al. found that the average ADC increases by 3% per
images acquired with the gradient pulses applied in three decade after the age of 40 [56]. This has been attributed to
orthogonal planes can be multiplied, and the cube root of that loss of myelinated neurons and structural changes seen
product yields a DW image weighted with diffusion magni- with aging.
tude information alone and directional information removed,
called the “isotropic DWI” (SDWI):
Current DWI Techniques
S DWI = ( S x S y S z )
1/ 3
(1.11)
The movement of water molecules detected by DWI occurs
The ADC derived from application of the gradients in
on a length scale of micrometers that is significantly larger
three orthogonal planes, called the “trace ADC” (simply
than intracellular distances, but much smaller than the milli-
called ADC below), is the average of the ADCs from each of
meter scale of voxel size in typical magnetic resonance (MR)
the planes (Fig. 1.3):
images. The original PGSE T2-weighted sequence described
ADC = ( ADC x + ADC y + ADCz ) / 3 (1.12) by Stejskal and Tanner was sensitive to even minimal bulk
patient motion, which was enough to obscure the much more
The images submitted to the radiologist for interpretation miniscule molecular motion of diffusion. As MRI technol-
in the clinical setting are typically the isotropic DWI and ogy has improved and high-performance gradients have been
trace ADC images. developed, DWI is now typically performed using spin-echo
single-shot echo-planar imaging (SS-EPI) techniques. With
this type of pulse sequence, an entire two-dimensional (2-D)
DWI of the Normal Brain image can be formed from a single RF excitation pulse
(hence the term “single-shot”). Images can be acquired in a
As mentioned previously, the average ADC in the brain is fraction of a second, minimizing artifacts related to patient
approximately 0.7 × 10−3 mm2/s; more specifically, mean dif- motion. Additionally, the SS-EPI technique has a relatively
fusivities in the adult brain range from 0.67 to 0.83 × 10−3 high signal-to-noise ratio (SNR); this is an important advan-
mm2/s in gray matter, and 0.64 to 0.71 × 10−3 mm2/s in white tage in DWI, as high b-value diffusion gradients cause a sig-
matter [10]. Because the ADC values for gray and white mat- nificant loss of signal intensity (see Eq. (1.4)). Using the
ter are so similar, there is essentially no contrast between SS-EPI technique, DW imaging of the brain can typically be
gray and white matter on the ADC map or exponential image. completed in 1–2 min, beneficial in clinical settings where
The subtle gray–white matter contrast that can be seen on the acutely ill and often uncooperative patients (i.e., hyperacute
DWI can be attributed to T2-effects, which makes it impor- stroke patients) are being imaged.
1 Physical Principles of Diffusion Imaging 11

Limitations of the SS-EPI technique include low spatial Conclusion

resolution, blurring effects of T2*- and T2-decay, and sen-
sitivity to artifacts. Matrix size using the single-shot tech- From the observation of Brownian motion in 1827 to the
nique is limited to 128 × 128 for a typical DWI using description of molecular diffusion by Einstein in 1905, and
current MRI hardware and software, compared to matrix the application of these principles to DW MRI, a great deal
sizes of 256 × 192 or larger for standard T1- and T2-weighted has been learned about the movement of water molecules in
sequences. Blurring of T2*- and T2-weighted contrast the brain (Table 1.1). DWI is a powerful tool that provides a
occurs during image readout due to the extended echo-train great deal of information about the underlying structure and
length. SS-EPI is also sensitive to artifacts due to Nyquist physiology of the brain that is not provided with standard
ghosting, chemical shift, and particularly magnetic field T1- and T2-weighted imaging techniques. DWI was first
inhomogeneities caused by local susceptibility differences
between adjacent structures. As stronger and faster gradi-
ents are developed to improve DWI, problems such as eddy Table 1.1 Summary of articles reviewing the basic principles of DWI
discussed in this chapter
currents and mechanical vibration may be exacerbated,
resulting in additional artifacts. These artifacts are Authors Topic Summary
Schaefer Diffusion-weighted Reviews the basic principles
explained in detail in reviews by Le Bihan et al. [59], and et al. MR imaging of the behind DW MR imaging of
by Mukherjee et al. [60]. (2000) [9] brain the brain, as well as clinical
The development of multichannel coils and parallel imag- applications of the technique
ing has also led to improvements in DWI. Compared to stan- Bammer Basic principles of Describes the physics behind
dard head RF coils, which have a uniform sensitivity (2003) [61] diffusion-weighted image contrast in DWI and
imaging various imaging sequences
throughout their imaging volume, the newer multichannel that can be utilized for DWI,
phased-array RF coils have increased sensitivity around their with respect to speed, spatial
periphery than at their center, resulting in improved SNR in resolution, and sensitivity to
the cerebral cortex and, therefore, improved DW images. motion; more advanced
diffusion measurement
These coils also allow parallel imaging on modern MR scan- techniques are also discussed
ners due to their multiple independent receiver channels. Hagmann Understanding Explains the physics of water
Parallel imaging techniques can be used to shorten the echo-­ et al. diffusion MR imaging diffusion and basic principles
train length of EPI, thereby reducing the susceptibility-­ (2006) [8] techniques: from of diffusion contrast encoding
scalar diffusion-­ with MRI; discusses and
induced and blurring artifacts that typically occur with weighted imaging to compares the spectrum of
extended echo trains. The shorter readout may also increase diffusion tensor diffusion-based MR imaging
SNR by decreasing the TE. imaging and beyond techniques, from the simplest
Alternatives to SS-EPI DWI include variations of fast to most complex
Le Bihan Artifacts and pitfalls Artifacts and pitfalls of
spin-echo (FSE) or turbo spin-echo (TSE) imaging, multi-
et al. in diffusion MRI diffusion-based MR imaging
shot EPI, spiral imaging, and line-scan imaging techniques (2006) [59] techniques are discussed, as
[60, 61]. SS-FSE and half-Fourier acquired single-shot turbo well as potential strategies to
spin-echo (HASTE) are ultrafast sequences that limit arti- overcome them
facts related to bulk patient motion similar to SS-EPI, but Jones Studying connections Review of the basic principles
(2008) [7] in the living human underlying DWI, and how
with reduced susceptibility and chemical shift artifacts com- brain with diffusion water diffusion can be utilized
pared to SS-EPI. However, they require longer scan times MRI to study underlying brain
due to their low SNR compared with SS-EPI, and therefore structure
have not become popular techniques for brain DWI imaging. Mukherjee Diffusion tensor MR Reviews technical
et al. imaging and fiber considerations in DWI,
Multishot techniques also confer a reduction in susceptibility (2008) [60] tractography: including current techniques
artifacts compared with SS-EPI, but scan times are longer, technical and their limitations,
which makes these methods more sensitive to bulk patient considerations optimization of current
motion. One multishot FSE technique reduces motion arti- techniques, and newer
techniques that have been
facts by continually oversampling the center of k-space, and developed for both DWI and
is called periodically rotated overlapping parallel lines with DTI
enhanced reconstruction (PROPELLER). This technique Mukherjee Diffusion tensor MR Review of the basic physics
improves detection of small acute infarcts, particularly at the et al. imaging and fiber and theory underlying DWI
(2008) [62] tractography: theoretic and ADC mapping, as well as
skull base and in the posterior fossa, where SS-EPI tech-
underpinnings DTI
niques have significant susceptibility-induced distortion.
ADC apparent diffusion coefficient, DW, diffusion-weighted, DWI
Due to much longer scanning times, however, the technique diffusion-­weighted imaging, DTI diffusion tensor imaging, MR mag-
has not surpassed SS-EPI for routine clinical DWI. netic resonance, MRI magnetic resonance imaging
12 T. Sivapatham and E. R. Melhem

routinely used in brain imaging for the detection of acute 19. Sevick RJ, Kanda F, Mintorovitch J, Arieff AI, Kucharczyk
ischemia, and is now an essential part of the imaging evalua- J, Tsuruda JS, et al. Cytotoxic brain edema: assess-
ment with diffusion-­ weighted MR imaging. Radiology.
tion of patients with acute stroke. The applications of DWI in 1992;185(3):687–90.
neuroimaging are extensive, and include the evaluation of 20. Ebisu T, Naruse S, Horikawa Y, Ueda S, Tanaka C, Uto M, Umeda
intracranial infections, neoplasms, demyelinating processes, M, Higuchi T. Discrimination between different types of white
and trauma. These will be discussed in the following chapter. matter edema with diffusion-weighted MR imaging. J Magn Reson
Imaging. 1993;3(6):863–8.
DWI techniques can also be used to evaluate the fiber tracts 21. Schaefer PW, Buonanno FS, Gonzalez RG, Schwamm
of the brain, and these techniques and applications are dis- LH. Diffusion-weighted imaging discriminates between cyto-
cussed elsewhere in this book. As MRI technology advances toxic and vasogenic edema in a patient with eclampsia. Stroke.
and improvements are made in imaging hardware and soft- 1997;28:1082–5.
22. Schwartz R, Mulkern R, Gudbjartsson H, Jolesz F. Diffusion-­
ware, newer techniques and sequences may provide more weighted MR imaging in hypertensive encephalopathy: clues to
sensitive DW images with reduction in the artifacts inherent pathogenesis. AJNR Am J Neuroradiol. 1998;19:859–62.
in current DWI techniques. 23. Ay H, Buonanno FS, Schaefer PW, Le DA, Wang B, Gonzalez
RG, Koroshetz WJ. Posterior leukoencephalopathy without severe
hypertension: utility of diffusion-weighted MRI. Neurology.
1998;51:1369–76.
References 24. Mukherjee P, McKinstry RC. Reversible posterior leukoencepha-
lopathy syndrome: evaluation with diffusion-tensor MR imaging.
1. Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L, Shimizu H, Radiology. 2001;219:756–65.
Kucharczyk J, et al. Early detection of regional cerebral ischemia 25. Provenzale JM, Petrella JR, Cruz LC Jr, Wong JC, Engelter S,
in cats: comparison of diffusion- and T2-weighted MRI and spec- Barboriak DP. Quantitative assessment of diffusion abnormali-
troscopy. Magn Reson Med. 1990;14:330–46. ties in posterior reversible encephalopathy syndrome. AJNR Am J
2. Chien D, Kwong KK, Gress DR, Buonanno FS, Buxton RB, Rosen Neuroradiol. 2001;22:1455–61.
BR. MR diffusion imaging of cerebral infarction in humans. AJNR 26. Chen PE, Simon JE, Hill MD, Sohn CH, Dickhoff P, Morrish WF,
Am J Neuroradiol. 1992;13:1097–102. et al. Acute ischemic stroke: accuracy of diffusion-weighted MR
3. Warach S, Chien D, Li W, Ronthal M, Edelman RR. Fast magnetic imaging–effects of b value and cerebrospinal fluid suppression.
resonance diffusion-weighted imaging of acute human stroke. Radiology. 2006;238(1):232–9.
Neurology. 1992;42:1717–23. 27. Kim HJ, Choi CG, Lee DH, Lee JH, Kim SJ, Suh DC. High-b-value
4. van Everdingen KJ, van der Grong J, Kappelle LJ, Ramos LM, diffusion-weighted MR imaging of hyperacute ischemic stroke at
Mali WP. Diffusion-weighted magnetic resonance imaging in acute 1.5T. AJNR Am J Neuroradiol. 2005;26(2):208–15.
stroke. Stroke. 1998;29(9):1783–90. 28. Le Bihan D. Looking into the functional architecture of the brain
5. Brown R. A brief account of microscopical observations made in with diffusion MRI. Nat Rev Neurosci. 2003;4(6):469–80.
the months of June, July and August 1827 on the particles contained 29. Cleveland GG, Chang DC, Hazlewood CF, Rorschach HE. Nuclear
in the pollen of plants, and on the general existence of active mol- magnetic resonance measurement of skeletal muscle: anisotropy
ecules in organic and inorganic bodies. Philos Mag. 1828;4:161. of the diffusion coefficient of the intracellular water. Biophys J.
6. Einstein A. Investigations on the theory of Brownian motion. 1976;16(9):1043–53.
New York: Dover; 1956. 30. Moseley ME, Cohen Y, Kucharczyk J, Mintorovitch J, Asgari HS,
7. Jones DK. Studying connections in the living human brain with Wendland MF, Tsuruda J, Norman D. Diffusion-weighted MR
diffusion MRI. Cortex. 2008;44(8):936–52. imaging of anisotropic water diffusion in cat central nervous sys-
8. Hagmann P, Jonasson L, Maeder P, Thiran JP, Wedeen VJ, Meuli tem. Radiology. 1990;176(2):439–45.
R. Understanding diffusion MR imaging techniques: from sca- 31. Moseley ME, Kucharczyk J, Asgari HS, Norman D. Anisotropy in
lar diffusion-weighted imaging to diffusion tensor imaging and diffusion-weighted MRI. Magn Reson Med. 1991;19(2):321–6.
beyond. Radiographics. 2006;26(Suppl 1):205–23. 32. Doran M, Hajnal JV, Van Bruggen N, King MD, Young IR, Bydder
9. Schaefer PW, Grant PE, Gonzalez RG. Diffusion-weighted MR GM. Normal and abnormal white matter tracts shown by MR imag-
imaging of the brain. Radiology. 2000;217(2):331–45. ing using directional diffusion weighted sequences. J Comput
10. Pierpaoli C, Jezzard P, Basser PJ, Barnett A, Di Chiro Assist Tomogr. 1990;14(6):865–73.
G. Diffusion tensor MR imaging of the human brain. Radiology. 33. Chenevert TL, Brunberg JA, Pipe JG. Anisotropic diffusion in
1996;201:637–48. human white matter: demonstration with MR techniques in vivo.
11. Bloch F. Nuclear induction. Phys Rev. 1946;70:460–74. Radiology. 1990;177(2):401–5.
12. Bloch F, Hansen WW, Packard M. Nuclear induction. Phys Rev. 34. Thomsen C, Henriksen O, Ring P. In vivo measurement of water
1946;69:127. self-diffusion in the human brain by magnetic resonance imaging.
13. Purcell EM, Torrey HC, Pound RV. Resonance absorption by Acta Radiol. 1987;28(3):353–61.
nuclear magnetic moments in a solid. Phys Rev. 1946;69:37. 35. Hong X, Dixon WT. Measuring diffusion in inhomogeneous sys-
14. Hahn EL. Spin echoes. Phys Rev. 1950;80:580–94. tems in imaging mode using antisymmetric sensitizing gradients. J
15. Carr HY, Purcell EM. Effects of diffusion on free precession in Magn Reson. 1992;99:561–70.
nuclear magnetic resonance experiments. Phys Rev. 1954;94:630–8. 36. Lian J, Williams DS, Lowe IJ. Magnetic resonance imaging in the
16. Torrey HC. Bloch equations with diffusion terms. Phys Rev. presence of background gradients and imaging of background gra-
1956;104:563–35. dients. J Magn Reson Ser. 1994;106:65–74.
17. Stejskal EO, Tanner JE. Spin diffusion measurements: spin echoes 37. Wimberger DM, Roberts TP, Barkovich AJ, Prayer LM, Moseley
in the presence of a time-dependent field gradient. J Chem Phys. ME, Kucharczyk J. Identification of “premyelination” by diffusion-­
1965;42:288–92. weighted MRI. J Comput Assist Tomogr. 1995;19(1):28–33.
18. Mori S, Barker PB. Diffusion magnetic resonance imaging: its prin- 38. Prayer D, Roberts T, Barkovich AJ, Prayer L, Kucharczyk J,
ciple and applications. Anat Rec. 1999;257(3):102–9. Moseley M, Arieff A. Diffusion-weighted MRI of myelina-
1 Physical Principles of Diffusion Imaging 13

tion in the rat brain following treatment with gonadal hormones. 50. Schneider JF, Il’yasov KA, Hennig J, Martin E. Fast quantitative
Neuroradiology. 1997;39(5):320–5. diffusion-tensor imaging of cerebral white matter from the neonatal
39. Beaulieu C, Allen PS. Determinants of anisotropic water diffusion period to adolescence. Neuroradiology. 2004;46(4):258–66.
in nerves. Magn Reson Med. 1994;31(4):394–400. 51. Snook L, Paulson LA, Roy D, Phillips L, Beaulieu C. Diffusion
40. Beaulieu C, Allen PS. Water diffusion in the giant axon of the tensor imaging of neurodevelopment in children and young adults.
squid: implications for diffusion-weighted MRI of the nervous sys- NeuroImage. 2005;26(4):1164–73.
tem. Magn Reson Med. 1994;32(5):579–83. 52. Hermoye L, Saint-Martin C, Cosnard G, Lee SK, Kim J, Nassogne
41. Beaulieu C, Allen PS. An in vitro evaluation of the effects of local MC, Menten R, Clapuyt P, Donohue PK, Hua K, Wakana S, Jiang
magnetic-susceptibility-induced gradients on anisotropic water dif- H, van Zijl PC, Mori S. Pediatric diffusion tensor imaging: normal
fusion in nerve. Magn Reson Med. 1996;36(1):39–44. database and observation of the white matter maturation in early
42. Beaulieu C. The basis of anisotropic water diffusion in the nervous childhood. NeuroImage. 2006;29(2):493–504.
system - a technical review. NMR Biomed. 2002;15(7):435–55. 53. Gideon P, Thomsen C, Henriksen O. Increased self-diffusion of brain
43. Burdette JH, Elster AD, Ricci PE. Acute cerebral infarction: quan- water in normal aging. J Magn Reson Imaging. 1994;4(2):185–8.
tification of spin-density and T2 shine-through phenomena on 54. Chun T, Filippi CG, Zimmerman RD, Uluğ AM. Diffusion
diffusion-­weighted MR images. Radiology. 1999;212(2):333–9. changes in the aging human brain. AJNR Am J Neuroradiol.
44. Hüppi PS, Maier SE, Peled S, Zientara GP, Barnes PD, Jolesz FA, 2000;21(6):1078–83.
Volpe JJ. Microstructural development of human newborn cerebral 55. Engelter ST, Provenzale JM, Petrella JR, DeLong DM, MacFall
white matter assessed in vivo by diffusion tensor magnetic reso- JR. The effect of aging on the apparent diffusion coefficient
nance imaging. Pediatr Res. 1998;44(4):584–90. of normal appearing white matter. AJR Am J Roentgenol.
45. Neil JJ, Shiran SI, McKinstry RC, Schefft GL, Snyder AZ, Almli 2000;175:425–30.
CR, Akbudak E, Aronovitz JA, Miller JP, Lee BC, Conturo 56. Chen ZG, Li TQ, Hindmarsh T. Diffusion tensor trace map-
TE. Normal brain in human newborns: apparent diffusion coeffi- ping in normal adult brain using single-shot EPI technique.
cient and diffusion anisotropy measured by using diffusion tensor A methodological study of the aging brain. Acta Radiol.
MR imaging. Radiology. 1998;209(1):57–66. 2001;42(5):447–58.
46. Mukherjee P, Miller JH, Shimony JS, Conturo TE, Lee BC, Almli 57. Nusbaum AO, Tang CY, Buchsbaum MS, Wei TC, Atlas
CR, McKinstry RC. Normal brain maturation during childhood: SW. Regional and global changes in cerebral diffusion with normal
developmental trends characterized with diffusion-tensor MR aging. AJNR Am J Neuroradiol. 2001;22(1):136–42.
imaging. Radiology. 2001;221(2):349–58. 58. Moseley M. Diffusion tensor imaging and aging - a review. NMR
47. Mukherjee P, McKinstry RC. Diffusion tensor imaging and trac- Biomed. 2002;15(7-8):553–60.
tography of human brain development. Neuroimaging Clin N Am. 59. Le Bihan D, Poupon C, Amadon A, Lethimonnier F. Artifacts and pit-
2006;16(1):19–43. falls in diffusion MRI. J Magn Reson Imaging. 2006;24(3):478–88.
48. Mukherjee P, Miller JH, Shimony JS, Philip JV, Nehra D, Snyder AZ, 60. Mukherjee P, Chung SW, Berman JI, Hess CP, Henry RG. Diffusion
Conturo TE, Neil JJ, McKinstry RC. Diffusion-tensor MR imaging tensor MR imaging and fiber tractography: technical consider-
of gray and white matter development during normal human brain ations. AJNR Am J Neuroradiol. 2008;29(5):843–52.
maturation. AJNR Am J Neuroradiol. 2002;23(9):1445–56. 61. Bammer R. Basic principles of diffusion-weighted imaging. Eur J
49. McGraw P, Liang L, Provenzale JM. Evaluation of normal age-­ Radiol. 2003;45(3):169–84.
related changes in anisotropy during infancy and childhood 62. Mukherjee P, Berman JI, Chung SW, Hess CP, Henry RG. Diffusion
as shown by diffusion tensor imaging. AJR Am J Roentgenol. tensor MR imaging and fiber tractography: theoretic underpinnings.
2002;179(6):1515–22. AJNR Am J Neuroradiol. 2008;29(4):632–41.
 Physical Principles of Dynamic
Contrast-­Enhanced and Dynamic 2
Susceptibility Contrast MRI

Mark S. Shiroishi, Jerrold L. Boxerman, C. Chad Quarles,

Daniel S. R. Stahl, Saulo Lacerda, Naira Muradyan,
Timothy P. L. Roberts, and Meng Law

Introduction passage of a GBCA through tissue decreases the intrinsic T1,

T2, and T2* relaxation times [2]. This chapter will provide an
The use of dynamic contrast-agent-enhanced magnetic reso- overview of the general physical principles of these techniques.
nance imaging (MRI) can provide insight into hemodynamic An overview of these two methods is provided in Table 2.1 [3].
processes not detectable using conventional contrast-­enhanced
magnetic resonance (MR) techniques. This additional data may
Table 2.1 Overview of DCE-MRI and DSC-MRI
allow refinement of differential diagnoses based on microvas-
DCE-MRI DSC-MRI
cular physiology. The dominant dynamic gadolinium-based
Bolus handling Bolus passage Bolus tracking
contrast agent (GBCA) injection MRI techniques currently uti- Acquisition Accumulation of contrast First-pass of contrast
lized in brain imaging are: (1) T1-weighted dynamic contrast- point agent agent
enhanced (DCE) MRI, and (2) T2/T2*-weighted dynamic Contrast media Intravenous bolus Intravenous bolus
susceptibility contrast (DSC) MRI. Of these, DSC-MRI is injection of a GBCA injection of a GBCA
much more commonly used for clinical perfusion imaging of Tracer Flow or permeability-­ Non-diffusible blood
limited diffusible tracer pool tracer
the brain, especially for the evaluation of stroke and tumor. On Relaxation T1 relaxation T2/T2* relaxation
the other hand, DCE-­MRI is the dominant method of dynamic mechanism
contrast-enhanced MRI outside of the brain [1]. In both DCE- Effect T1 shortening effect Increased susceptibility
MRI and DSC-­MRI, dynamic images are acquired before, dur- effect
ing, and after the administration of an exogenous GBCA. As Signal Increased signal Decreased signal
behaviors
opposed to other techniques, such as contrast-enhanced com-
DCE-MRI dynamic contrast-enhanced magnetic resonance imaging,
puted tomography (CT), contrast-enhanced MRI is distinctive
DSC-MRI dynamic susceptibility contrast magnetic resonance imaging,
because it detects the changes induced in the local relaxation GBCA gadolinium-based contrast agent
times of water rather than detecting the GBCA itself, where the Source: Adapted under terms of Creative Commons license from [3]

M. S. Shiroishi (*) D. S. R. Stahl

Division of Neuroradiology, Department of Radiology, Keck Rutgers New Jersey Medical School, Newark, NJ, USA
School of Medicine of University of Southern California,
S. Lacerda
Los Angeles, CA, USA
Radiologia - Neurorradiologia na Hospital da Bahia, Salvador,
e-mail: [email protected]
Bahia, Brazil
J. L. Boxerman
N. Muradyan
Diagnostic Imaging, Warren Alpert Medical School of Brown
U.S. Food and Drug Administration, Center for Devices and
University, Providence, RI, USA
Radiological Health, Silver Spring, MD, USA
Department of Diagnostic Imaging, Rhode Island Hospital, e-mail: [email protected]
Providence, RI, USA
T. P. L. Roberts
e-mail: [email protected];
Department of Radiology, Children’s Hospital of Philadelphia,
[email protected]
Philadelphia, PA, USA
C. C. Quarles e-mail: [email protected]
Division of Neuroimaging Research, Barrow Neurological
M. Law
Institute, St. Joseph’s Hospital and Medical Center,
Nuclear Medicine, Neuroscience, Electrical and Computer
Phoenix, AZ, USA
Systems Engineering, Monash University, Alfred Health
e-mail: [email protected]
Organization, Melbourne, VIC, Australia
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 15

S. H. Faro, F. B. Mohamed (eds.), Functional Neuroradiology, https://doi.org/10.1007/978-3-031-10909-6_2
16 M. S. Shiroishi et al.

T1-Weighted Dynamic Table 2.2 QIBA DCE-MRI acquisition parameters for brain imaging
Contrast-Enhanced MRI Parameter DCE-MRI
Field strength 1.5/3T
When applied to the brain, DCE-MRI is primarily employed Acquisition sequence 3D SPGR
to characterize the functional integrity of the blood–brain Receive coil type ≥8 channel head array coil
Lipid suppression On
barrier (BBB) via estimation of microvascular permeability
Slice thickness ≤5 mm
to GBCAs. The evaluation of cancer is a major application of Gap thickness 0–1 mm
DCE-MRI where it has potential to provide prognostic, pre- FOV 220–240 mm
dictive, and physiological response imaging biomarkers. Acquisition matrix 256 × 128–160
Conventional GBCAs used in clinical MRI are diffusible, Plane orientation Axial
low-molecular-weight extracellular agents (~500 to 1000 Phase/frequency encode direction AP/RL
Da) that remain intravascular when the BBB is intact. Receiver bandwidth 250 Hz/pixel
Pre-contrast Post-contrast
Disruption of the BBB secondary to a variety of pathological
# Phases ≥5 40–80
processes results in the transfer of GBCA moieties across the
# Averages ≥1 1
capillary endothelium from the intravascular space into the Flip angles 2–30a 25–30
extravascular–extracellular space (EES). Leakage of GBCAs TR (ms) 3–8 msb 3–8
into the EES results in T1-shortening and contrast enhance- TE ≤3 msb ≤3 ms
ment on T1-weighted imaging. Temporal resolution <10 (ideal 5) s
Total acquisition time 5–10 min
3D three dimensional; DCE-MRI dynamic contrast-enhanced magnetic
DCE-MRI Acquisition resonance imaging, FA flip angle, FOV field of view, SPGR spoiled gra-
dient recalled aqcquisition; TE echo time, TR repetition time
Source: Adapted with permission from [5]
There has historically been quite a variation of DCE-MRI a
Variable FAs for T10 measurement
acquisition protocols in the literature. DCE-MRI acquisition b
Ensure that TR/TE stays constant for all FAs
parameters are generally intended to emphasize R1 contrast
and minimize competing T2* effects by employing short with parallel imaging (GRAPPA) along with two-dimensional
echo times (TEs) and repetition times (TRs) [2]. DCE-MRI (2D) or three-­ dimensional (3D) + simultaneous multislice
most often utilizes a fast T1-weighted spoiled gradient-­ imaging (SMS) encoding [9], radial k-space encoding with
recalled echo sequence with the temporal resolution contin- golden angle ordering (GRASP) method [10], high under-sam-
gent on the volume coverage, contrast-to-noise ratio (CNR), pling factors, [11] and time-resolved MR angiography methods
and spatial resolution for a particular organ system [4, 5]. with keyhole view-sharing [12].
The temporal resolution demands for DCE-MRI are gener-
ally less than that for DSC-MRI unless an arterial input func-
tion (AIF) is needed [2]. DCE-MRI scan durations are DCE-MRI Data Analysis
generally much longer than for DSC-MRI and to estimate
microvascular permeability with DCE-MRI, the temporal In DCE-MRI, the concentration of GBCA must be deter-
resolution generally ranges between 5 and 20 s [6–8]. Like mined in order to perform pharmacokinetic (PK) modeling.
with DSC-MRI, consistent technique including the use of a This is accomplished by measuring changes in T1-weighted
power injector for bolus injection (2–4 cc/s) of GBCA fol- signal intensity and assuming that these changes are propor-
lowed by a 20–30 cc saline flush at the same rate into the tional to GBCA concentration. This is a commonly used
right arm to decrease possible venous reflux should be per- assumption given its simplicity; however, at high tracer con-
formed if possible in all cases. centrations, the relationship between signal intensity and
Recent initiatives such as those by the Radiological Society GBCA concentration is non-linear and this can result in sys-
of North America’s (RSNA’s) Quantitative Imaging Biomarkers tematic error of DCE-MRI parameters.
Alliance (QIBA) have focused on standardizing acquisition Some studies have utilized predetermined T1 values,
and analysis of various imaging methods including DCE- usually from the literature. However, this too can result in
MRI. QIBA recommendations for DCE-MRI acquisition are bias for several reasons: these are often performed in healthy
included in Table 2.2 [5]. (Please see section “Standardization subjects and there is then no consideration of the effects of
Efforts and Variability of DCE-MRI” below). In addition to aging or pathological conditions nor does it consider indi-
conventional DCE-MRI acquisition methods, there have been vidual variability or tissue heterogeneity [1, 13–18].
several recent advances in pulse sequence acceleration meth- Therefore, direct measurement of T1 in a given individual is
ods to obtain high temporal and/or spatial resolution in DCE- more desirable because of potential T1 variability, particu-
MRI. These include dynamic compressed sensing combined larly from pathological states. The gold standard method of
2 Physical Principles of Dynamic Contrast-Enhanced and Dynamic Susceptibility Contrast MRI 17

T1 mapping uses inversion recovery (IR) sequences; how- kinetics of GBCA tissue accumulation [28, 29]. While easier
ever, their long acquisition times preclude routine use [19]. to perform than PK modeling of DCE-MRI data, these semi-­
The most commonly employed form of clinical T1 mapping quantitative methods cannot distinguish between physiologic
is the variable flip angle (VFA) method using flip angles factors and physical properties of image acquisition includ-
(FAs) of 2–30° with a gradient echo sequence (Table 2.2) ing, but not limited to, scanner parameters, method of GBCA
[5]. The accuracy of T1 mapping relies on FA accuracy, administration, and native T1 of the interrogated tissue
which can be compromised due to several factors including [30–32].
the presence of standing waves from dielectric resonance in
a subject [20], less uniform FA from smaller transmit coils,
and poor slice profiles from 2D multi-slice imaging [21]. B1 DCE Pharmacokinetic Modeling
mapping can be particularly helpful to correct FA inaccura-
cies, particularly at 3T and over large anatomic coverage [5, Through more sophisticated PK modeling of DCE-MRI
22]. In addition to the IR and VFA methods, the Look- data, various quantitative parameters can be determined: the
Locker (LL) method is another technique for T1 mapping. volume transfer constant between blood plasma and the EES
Like the VFA method, LL methods are faster than IR; how- (Ktrans), the volume of EES per unit volume of tissue (ve), the
ever, they may also result in errors with studies suggesting rate constant between EES and blood plasma (kep, where kep
that VFA may result in overestimation while LL may under- = Ktrans/ve), capillary wall permeability surface area product
estimate T1 values due to inaccurate B1 mapping and per unit volume of tissue (PSρ), and capillary blood flow
incomplete spoiling [19]. (perfusion) per unit volume of tissue (Fρ) [6]. Most DCE-­
The determination of the arterial input function (AIF) to MRI tracer kinetic models divide the tissue of interest into
obtain more accurate measurement of the concentration of several compartments (Fig. 2.1). These include the blood
GBCA in blood plasma (Cp(t)) can be another source of error plasma volume per unit volume of tissue (vp) and the volume
in DCE-MRI. Partial-volume average artifacts due to limita- of extravascular–extracellular space per unit volume of tis-
tions in spatial resolution can result from the sampling of a sue (ve).
small cerebral artery. Sampling from a large vein such as the Of the various PK models that have been used to analyze
superior sagittal sinus as a venous outflow function (VOF) DCE-MRI data, the most popular model is commonly
can be used to correct partial-volume average artifacts by referred to as the Tofts model [6, 33] and provides measures
rescaling the area under the AIF curve [23–25]. The use of of Ktrans and ve. Ktrans was originally described by the follow-
three-dimensional (3D) image acquisition, ensuring that the ing equation:
artery of interest is well visualized in the excitation slab, or
dCe ( t )
using non-selective saturation pre-pulse can help alleviate ve = K trans ( C p ( t ) − Ce ( t ) )
inflow artifacts where the arterial blood appears bright on dt
pre-contrast images [26]. If measurement of an individual where Cp and Ce are the blood plasma and EES contrast agent
AIF is not practical, other alternatives that have been used concentrations, respectively (Fig. 2.1). Ktrans is the most fre-
include population-based AIFs that do not incorporate indi- quently utilized metric in DCE-MRI and describes the rate of
vidual differences. There are also reference tissue models contrast agent flux into the EES. Its physiological meaning
that attempt to estimate the vascular tracer concentration
from one or more normal-appearing surrounding tissues
[27]. Hematocrit values should theoretically be incorporated
into the AIF measurement because the GBCA remains in the
blood plasma component and does not pass into red blood
cells. However, in practice, a standard, rather than a directly
measured, hematocrit value is used and this too can result in
errors [4].
There are a variety of methods to analyze DCE-MRI data.
At its most basic, non-PK modeling methods use subjective
assessment of the signal intensity-time curve. These are sim-
ple to perform and interpret, yet will not provide in-depth
understanding of the underlying pathophysiology [28]. Other Fig. 2.1 Schematic of two-compartment model in DCE-MRI. Keys: vp
methods involve semi-quantitative analysis of data with met- = blood plasma volume per unit volume of tissue; ve = volume of extra-
rics such as the initial area under the enhancement curve vascular extracellular space per unit volume of tissue; Ktrans = volume
transfer constant between blood plasma and EES; Cp = tracer concen-
(IAUC) and other methods of signal intensity-time curve tration in arterial blood plasma; Ce = tracer concentration in EES; and
analysis that provide more detailed characterization of the blue circles = intracellular space where contrast agent is excluded
18 M. S. Shiroishi et al.

can be complex as it is dependent on vascular permeability, where vp represents the blood plasma volume per unit vol-
capillary surface area, and the type of contrast agent utilized ume of tissue. This model is often referred to as the “extended
[34]. When there is very high permeability of the endothe- Tofts model” [6] (Fig. 2.1). vp may be ignored in situations
lium with respect to blood flow, (F << PS), Ktrans primarily when the plasma volume or tracer concentration is negligi-
reflects blood flow—(Ktrans = F(1 − Hct)—as this is the main ble; e.g., hypovascular low-enhancing tissues or a few min-
limiting factor of the contrast agent flux; in this case, DCE-­ utes after the bolus. However, in diseases that are highly
MRI could be seen as “perfusion imaging.” When there is perfused, such as neoplasms, there should be consideration
very low permeability as compared to blood flow (F >> PS), of vp [4] (Fig. 2.2).
Ktrans mainly reflects permeability (Ktrans = PS), and in these There are less commonly utilized PK models besides the
situations DCE-MRI could be referred to as “permeability Tofts and extended Tofts models. While the Tofts models
imaging” [6, 34, 35]. assume a bi-directional exchange of CA between the vascu-
In the original Tofts model, neglecting the contribution of lar space and EES, a simpler assumption of a unidirectional
intravascular tracer to the MRI signal may be appropriate for transport of CA from the vascular to the EES compartment
a diffusible tracer where its distribution volume is large rela- can be formulated. The “Patlak model” [37] utilizes this
tive to blood volume. However, with an extracellular tracer, form and can be expressed as:
this may be problematic as its distribution volume is smaller t
[6, 36]. This assumption may produce erroneous Ktrans esti- Ct ( t ) = v p C p ( t ) + K trans ∫C p (τ ) dτ
mates because intravascular tracer could contribute a signifi- 0

cant proportion of the observed tissue signal. Therefore, in The two-compartment exchange model (2XCM) is a
the presence of an intravascular–extracellular tracer, the more generalized kinetic model than the Tofts and Patlak
model has been modified and expressed as: models. It can be used in mixed perfusion and permeability
t conditions that can allow estimation of PS and F to be calcu-
Ct ( t ) = v p C p ( t ) + K trans ∫C p (τ ) e
− K trans ( t −τ ) / ve
dτ lated [1, 38, 39]. This takes the form of:
0

dC p ( t ) dCe ( t )
vp = F ( Ca − C p ) − K PS ( C p − Ce ) and ve = K PS ( C p − Ce )
dt dt

where Ca represents the AIF and KPS now can be thought of increased concern regarding sources of error during data
as Ktrans without the F versus PS uncertainty. It is important to acquisition and analysis [40, 41].
note that more complex modeling such as this necessitates

a b c

Fig. 2.2 A 52-year-old female with pathology-proven high-grade gli- aspect of the right temporal lobe. (c) Ktrans color map demonstrating a
oma. (a) Axial contrast-enhanced T1-weighted image and (b) axial lesion with high values in the enhancing wall of the tumor
T2-weighted image demonstrate an enhancing tumor in the medial
2 Physical Principles of Dynamic Contrast-Enhanced and Dynamic Susceptibility Contrast MRI 19

 tandardization Efforts and Variability

S ticularly important in longitudinal DCE-MRI studies. In
of DCE-MRI order to analyze image data from the QIBA DCE-MRI
phantom, QIBA also provides automated T1 quantification
Lately, there has been increasing awareness regarding the software [45].
need to decrease bias and variability of quantitative imaging A recent multicenter phantom study of vendor-provided
biomarkers. Efforts such as the RSNA’s QIBA have focused B1 mapping sequences demonstrated the potential for these
on various imaging methods including DCE-MRI [42]. The techniques to provide unbiased and reproducible quantifica-
QIBA Perfusion Biomarker Committee Task Force has been tion of B1 field inhomogeneity that could be used to account
continuously working on their DCE-MRI profile [43]. The for spatial variation in the transmitted radio frequency (RF)
goal of QIBA profiles such as the one on DCE-MRI is to field [46]. Version 2.0 of the RSNA QIBA DCE-MRI Profile
assist in achieving adequate performance for an imaging bio- is currently being written and it will address spatially depen-
marker and to provide details about the capabilities and limi- dent B1 field inhomogeneity effects that may affect VFA T1
tations of an imaging marker. It does so by offering guidance measurements. This is particularly problematic at higher
regarding imaging acquisition, devices, technologists, radi- fields like 3T and when data are acquired over large ana-
ologists, subject handling, image quality assurance, recon- tomic regions and may necessitate B1 mapping and correc-
struction software, imaging analysis tools, and image quality tions to be incorporated in the measurement of T1 values [5].
assurance. While the QIBA T1 phantom is a static phantom, there
Standardization of image acquisition parameters is a has been recent work by Kim et al. [47] on a dynamic perfu-
major point of emphasis for QIBA. Inter-scanner and inter-­ sion phantom focused on DCE-MRI of the abdomen. They
site variability of T1 values in the brain is well known where used two different 3T MRI scanners and three healthy volun-
the Look-Locker IR method can underestimate while the teers. When compared to a static phantom, they found that
VFA technique can overestimate white matter T1 measure- the perfusion phantom significantly decreased the variability
ments [19]. Factors such as the particular MR sequence of contrast concentration and Ktrans measurements measured
employed, B1 field inhomogeneity, temperature of the mag- in four abdominal organs (liver, spleen, pancreas, and para-
net bore, and incomplete spoiling of transverse magnetiza- vertebral muscles). One should note that while estimates of
tion can influence the derived T1 values [44]. Before the DCE-MRI performance can be conducted with phantoms,
acquisition of clinical DCE-MRI data, it is important to these experiments likely underestimate the variability pro-
determine the true scanner variance and bias for T1 values duced in clinical populations due to the absence of motion
through the use of a T1 phantom. The QIBA DCE-MRI T1 artifacts [5].
phantom is composed of spheres containing solutions of Few clinical DCE-MRI studies of variability have been
varying concentrations of nickel chloride [44]. The phantom done in the brain and more data are desperately needed.
contains two sets of spheres: one set to simulate the vascular However, practical difficulties centering on the need to do
input function and the other set to represent tissue (Fig. 2.3). multiple GBCA injections in patients make such studies dif-
The T1 values for the vascular input spheres range between ficult to conduct. One such study was performed in 2003 by
0.75 and 41.6 s−1 while the tissue spheres range between Jackson et al. [48] in 9 glioma patients and found that the
0.67 and 7.5 s−1. The phantom is filled with 30-mM sodium within-region of interest (ROI) coefficient of variation for
chloride solution to simulate patient coil loading. To obtain mean Ktrans was 7.7% with a repeatability coefficient of
T1 values, an acquisition protocol that encompasses the 21.3%. A more recent publication by Barboriak et al. in 2019
typical VFAs is used for T1 mapping. This employs a coro- [49] found that in a multicenter imaging study of recurrent
nal fast spoiled gradient echo sequence with VFAs of 2, 5, glioblastoma, less variation in inter-reader tumor segmenta-
10, 15, 20, 25, and 30°. The use of the QIBA DCE phantom tion volumes, possibly through the use of automated tools,
to determine test–retest reliability and T1 accuracy is par- may decrease variability in DCE-MRI metrics like Ktrans.
20 M. S. Shiroishi et al.

a b

Loc: -6.963
IM: 12
Series: 2
FOV: 360 mm
TR/TE/FA/NEX: 4.7/1.3/30/4

c VIF d Tissue
45 12

40
10
35

30 8
Measured R1 (S1)

Measured R1 (S1)

25
6
20

15 4

10
2
5

0 0
0 5 10 15 20 25 30 35 40 45 0 2 4 6 8 10 12
NIST Theoretical R 1 (S-1) NIST Theoretical R 1 (S-1)

Fig. 2.3 (a) The QIBA dynamic contrast-enhanced phantom layout Standards and Technology (NIST) theoretical R1 values. (d) R1 values
with 32 spheres, with different concentrations of NiCl2 solutions for for the 24 tissue-mimicking inserts. (Images contributed by Edward
varying T1 relaxation rates (R1). (b) T1-weighted MR image of the Jackson, University of Wisconsin-Madison. Reprinted with permission
phantom showing the 32 spheres and (c) R1 values of the eight-vascular from [5]
input function-mimicking inserts compared with National Institute of

 2/T2*-Weighted Dynamic Susceptibility

T Also sometimes referred to as bolus tracking MRI, a non-­
Contrast MRI diffusible tracer, typically a GBCA, is administered and
rapid images are obtained during the first-pass of the contrast
Dynamic susceptibility contrast (DSC)-MRI has been agent. Several parameters are derived from DSC-MRI
applied to many neurological diseases, most prominently including relative cerebral blood flow (CBF), mean transit
brain tumors and stroke. Compared to DCE-MRI, DSC-MRI time (MTT), and relative cerebral blood volume (rCBV).
is much more commonly used in the clinical setting for brain rCBV is generally considered the most widely utilized and
imaging, though the opposite is true outside of the brain. robust DSC-MRI perfusion metric in brain imaging.
2 Physical Principles of Dynamic Contrast-Enhanced and Dynamic Susceptibility Contrast MRI 21

DSC-MRI Acquisition acquisition of GRE and SE data without adding additional

scan time or GBCA injections. These methods may provide
Like DCE-MRI, DSC-MRI acquisition consists of images important additional complementary information given their
acquired before, during, and after administration of an intra- different sensitivities to vessel size, and enable vessel size
vascular contrast administration (CA). While DCE-MRI imaging (VSI) to be performed and also provide simultane-
emphasizes T1 contrast and uses short TE and TR to mini- ous DCE-MRI metrics with only a single dose of contrast
mize competing T2* effects, DSC-MRI emphasizes T2* and agent. Other techniques such as spiral perfusion imaging
T2 contrast. Accordingly, long TE and TR are used to mini- with consecutive echoes (SPICE) methods [53] can also pro-
mize competing T1 effects [2]. In gradient echo-echo planar vide both DSC-MRI and DCE-MRI metrics in a single
imaging (GRE-EPI DSC-­MRI), TE is usually in the range of acquisition and without the need of a preload dose of GBCA.
25–35 ms in order to optimize T2* weighting, signal-to-noise The scan duration of DSC-MRI is much shorter than
ratio (SNR), and sensitivity to T1 effects [2, 50, 51]. With DCE-MRI and is shortest for indications like brain tumor
regard to TR, 1.5 s or less is recommended to optimize tem- evaluation (at least 2 min recommended). For other indica-
poral resolution given the constraints of desired slices, TE tions where there will be bolus dispersion and delay, like
and T1 weighting [2, 50, 51]. Regardless of whether GRE or stroke evaluation, longer scan duration is needed [66]. DSC-­
spin echo (SE) sequences are used, DSC-MRI requires very MRI acquisitions are limited by compromises in spatiotem-
robust temporal resolution (<2 s/time point) with adequate poral resolution, volume coverage, and SNR. Recent
spatial resolution. Given the need for high temporal resolu- advances to accelerate DSC-MRI acquisition and achieve
tion, single-shot EPI sequences are typically used. Some optimal spatiotemporal resolution include the use of parallel
newer methods designed to address the image distortion and imaging methods [55, 67] to decrease the EPI readouts,
signal dropout artifacts that degrade traditional EPI acquisi- reduce EPI-related artifacts, and decrease partial-volume
tions include single-line acquisitions [27], spiral or radial effects. Other methods such as simultaneous multi-slice
acquisitions [52, 53], and advanced EPI readouts [54, 55]. acquisitions [68] can accelerate DSC-MRI acquisitions by
A flip angle of 60–70° may in principle provide a compro- applying simultaneously exciting multiple slice planes with
mise between SNR and T1 sensitivity from GBCA leakage radiofrequency pulses without significant loss of SNR while
effects [2, 50, 51]. Higher flip angles would result in greater achieving high spatiotemporal resolution.
SNR, but would be more susceptible to GBCA leakage Compared to DCE-MRI, GBCA injection rates should be
effects due to increased T1 sensitivity. On the other hand, relatively higher (at least 4 cc/s) for DSC-MRI in order to
lower flip angles are less prone to GBCA leakage effects but avoid underestimation of DSC-MRI metrics from slower
at the cost of lower SNR. In order to avoid partial-volume rates [69]. Similar to DCE-MRI, injection should ideally be
average artifacts, adequate spatial resolution of 1–3 mm in given via the right arm in order to avoid venous reflux.
plane and 3–5 mm through plane are recommended, though Approximately 60 s of baseline data should be acquired prior
this may depend on desired temporal resolution [2]. to the injection of a GBCA in order to provide good CBV
Most DSC-MRI data are acquired using GRE-EPI map CNR [51]. An overview of DSC-MRI recommended
sequences, although some have used SE-EPI. SE methods acquisition parameters is given in Table 2.3 [51, 70, 71].
are most sensitive to smaller vessels (<20 μ[mu]m; i.e., cap-
illaries) [56], and also are less prone to artifacts at bone–
brain–air interfaces or at the skull base compared to GRE DSC-MRI Data Analysis
methods [57, 58]. GRE-EPI methods result in greater signal
loss [2], and are sensitive to vessels of all sizes [56] with Based on the indicator dilution methods for non-diffusible
excellent signal-to-noise ratios. GRE methods also allow tracers, CBV is proportional to the area under the contrast
higher temporal resolution due to shorter TEs and this agent concentration (Δ[Delta]R2* [or Δ{Delta}R2])-time
improves AIF quantification [2]. Given its sensitivity to ves- curve, assuming that there is no contrast agent leakage or
sels of all sizes, perfusion metrics derived from GRE could recirculation [72]. While in DCE-MRI, dipole–dipole inter-
suffer from large vessel blooming artifact because of a ten- actions are primarily responsible for GBCA-based T1 relax-
dency of the signal from capillaries to be dominated by mac- ation enhancement, the main contrast mechanism in
rovascular signal [59, 60]. Thus, GRE-based acquisition DSC-MRI is susceptibility effects induced by GBCAs [73].
could provide overestimates of perfusion metrics, while SE When a GBCA is given as a bolus, a transient drop in signal
techniques likely provide truer estimates of capillary-level intensity is seen on the signal intensity-time curve, known as
perfusion compared to positron emission tomography (PET). “negative enhancement,” as opposed to the “positive
Newer multi-echo DSC-MRI acquisitions such as spin enhancement” due to enhanced T1 relaxation in DCE-MRI
and gradient echo (SAGE) [61–65] allow simultaneous or conventional contrast-enhanced T1-weighted imaging.
22 M. S. Shiroishi et al.

Table 2.3 DSC-MRI acquisition parameters for brain imaging between ΔR2∗(t) and C(t) may not hold true [75] and assump-
Parameter DSC-MRI tion of a quadratic relationship could be assumed to mitigate
Field strength 1.5/3T CBF errors, particularly when estimating the AIF [76].
Acquisition sequence Generally GRE-EPI rather than The area beneath the concentration-time curve is calcu-
SE-EPI lated to derive the CBV map. By applying tracer kinetic
Slice thickness 3–5 mm
modeling for intravascular tracer agents [77–80], CBV can
FOV 200 × 200 mm (range, 200 × 200 to
240 × 240 mm) be obtained by integrating C(t) using the following
Acquisition matrix 128 × 128 (range, 64 × 64 to 256 × relationship:
256) t
GBCA injection rate At least 4 mL/s H f ∫ Ct ( t ) dt
TR 1.0–1.5 s (SE-EPI); minimum (vs “as CBV = t
0

short as possible”) for GRE-EPI; ρ ∫ Ca ( t ) dt

0
generally 1.0–1.5 s
TE 45 ms at 1.5T/30 ms at 3 T where Hf represents the difference in hematocrit between the
Flip angle 60°, newer consensus protocol AIF and capillaries, Ct(t) represents the concentration of
suggests 30° with no-preload GBCA
may perform as well as 60° with
GBCA in the tissues, ρ represents the brain tissue density,
single-dose preload at 3 T and Ca(t) represents the AIF. The AIF can be ignored because
Temporal coverage 120 time points of constraints in quantification due to limited temporal and
Preload GBCA dose Single dose (0.1 mmol/kg Gd), given spatial resolution and so “relative” CBV is commonly
(particularly for studies 5–10 min prior to dynamic imaging reported.
performed with a high flip
angle)
Indicator dilution theory can be used to model Ct(t) using
Baseline acquisitions prior At least 30–50 the following equation:
to GBCA injection t
Total acquisition time At least 2 min (brain tumors) Ct ( t ) = CBF ⋅ Ca ( t ) ⊗ R ( t ) = CBF ⋅ ∫Ca ( t ) R ( t − τ ) dτ
DSC-MRI dynamic susceptibility contrast magnetic resonance imaging, 0

GBCA gadolinium-based contrast agent, GRE-EPI gradient echo-echo where ⊗ represents convolution of Ca(t) and the tissue resi-
planar imaging, FOV field of view, SE-EPI spin echo-echo planar imag-
ing, TE echo time, TR repetition time due function R(t), which represents the amount of contrast
Source: Adapted with permission from [51, 70, 71] agent that remains in the tissue at time t.
The deconvolution of the Ca(t) and Ct(t) is needed to
quantify CBF. Of the various methods available, the most
Like in DCE-MRI, changes in DSC-MRI signal intensity are
commonly used model-independent method is singular value
converted to the tissue concentration of GBCA at time t
decomposition (SVD) [81] that is expressed as:
(C(t)). For DSC-MRI, this relation is noted on a voxel-wise
t j
Ct ( t ) = CBF ⋅ ∫Ca ( t ) R ( t − τ ) dτ ≈ ∆t ∑Ca ( ti ) R ( t j − ti )
basis as:
k  S (t )  0 i =0
C ( t ) ∝ k ⋅ ∆R2∗ ( t ) = − ln  
TE  S0  where it is assumed that R(t) and Ca(t) remain constant over
small time intervals and that cerebral and arterial concentra-
where k represents a proportionality constant (often set to tions are measured at equally spaced time points. In order
unity as it is not known a priori) that is dependent on tissue for SVD to determine R(t), methods have been devised to
type, field strength, contrast agent, and pulse sequence; decrease errors from the potential delay between the AIF
ΔR2∗(t) represents the change in the T2* relaxation rate at and tissue concentration curves [82] and to avoid physiolog-
time t; TE is the echo time; S(t) represents the signal inten- ically unreasonable results due to noise that lead to unstable
sity at time t; and S0 represents the baseline signal intensity solutions. The most common ways to address these two
before arrival of the GBCA. It is assumed that T1 effects issues are to implement a block-circulant AIF discretization
due to an intact BBB are not significant during DSC-MRI matrix with a truncated SVD regularization approach,
acquisition and that there is a linear relationship between respectively [2].
ΔR2∗(t) and C(t) [74]. However, in lesions such as brain Application of the central volume theorem allows calcu-
tumors, disruption of the BBB is common and necessitates lation of the mean transit time (MTT):
changes in acquisition and post-processing methods in
order to compensate for GBCA leakage effects (discussed CBV
MTT =
below). Furthermore, the assumed linear relationship CBF
2 Physical Principles of Dynamic Contrast-Enhanced and Dynamic Susceptibility Contrast MRI 23

Aside from CBV, CBF, and MTT, there are several other agent leakage correction, and AIF considerations [2, 50, 51].
emerging DSC-MRI parameters on the horizon. Newer As a result, most DSC-MRI studies rely on qualitative or
kinetic models can provide estimates of oxygen extraction semi-quantitative measures. Most often, a summary statistic
fraction (OEF) and capillary transit time heterogeneity in the form of “relative” CBV (rCBV) or CBF (rCBF) is
(CTH) that may better relate to the oxygen delivery that often used without definition of the AIF [50]. It should be
could be attained for a given CBF [83]. Traditional determi- noted that in addition to “relative” CBV (or CBF), “rCBV”
nation of oxygen availability in the brain is determined using can also refer to “regional” CBV [99]. It is also common for
CBF and arterial oxygen concentration, but these newer relative CBV to refer to a value that is normalized to “nor-
methods have the potential to highlight perfusion derange- mal” tissue, typically contralateral white matter [100], while
ments in brain tissue that may not be detected with conven- “regional” CBV often refers to absolute quantification of
tional DSC-MRI analysis. As was mentioned previously, the CBV.
use of gradient-echo and spin-echo sequences to provide Scaling metrics such as normalization or standardization
simultaneous estimations of Δ(Delta)R2 and Δ(Delta)R2* are commonly applied to non-quantitative rCBV values in
can provide other parameters including measures of vessel order to compare between subjects and imaging sessions.
size imaging (VSI), microvascular density, mean vessel However, the amount of variability intrinsic to these tech-
diameter [84], and vessel architectural imaging (VAI) niques is unclear [101]. With normalization, the mean value
(Fig. 2.4) [85]. In current practice, these measures are of the voxels within a tumoral ROI is divided by those in a
obtained with tissue sampling and defined by the pathologist. reference ROI, usually that in normal-appearing white mat-
However, validation of these techniques could overcome the ter. Normalization is quite commonly used; however, it can
limitations of sampling error and inability to perform longi- be time consuming and lead to user-dependent subjectivity
tudinal analysis, and may become important with the contin- [102]. On the other hand, when standardization is used, there
ued development of anti-vascular and anti-angiogenic is no need to use a reference ROI because rCBV maps are
therapies [86]. transformed to a standardized intensity scale. In this way, it
can function as an objective technique of converting rCBV
values to a consistent scale and it appears to improve rCBV
Arterial Input Function measurement consistency across patients and time [102].

Determination of the AIF is one of the leading sources of

error in the quantification of DSC-MRI. Various manual and Leakage Effects of GBCAs
automatic approaches [87–94] have been proposed to mea-
sure the AIF, but the most commonly used approach is to use GBCA leakage can diminish the accuracy and precision of
a global, as opposed to local, measurement using voxels rCBV derived from DSC-MRI. With an intact BBB, com-
either in or adjacent to the middle cerebral artery (MCA). partmentalization of GBCA within the vasculature mainly
While straightforward to do, assuming a global AIF will not impacts T2 or T2* with minimal impact on T1, and dimin-
likely be the true arterial input to the region of interest and utive Δ(Delta)R1 is a major assumption in DSC-MRI [56].
may introduce errors in quantification due to AIF delay and In theory, the equation for Δ(Delta)R2* is valid only if the
dispersion [66, 95]. Delay effects can be compensated for by changes in T1 associated with GBCA leakage do not sig-
using AIF discretization techniques that are resistant to delay nificantly affect signal intensity. However, this assumption
in SVD for example, while dispersion effects are difficult to often does not hold true as a disrupted blood–brain barrier
adjust for when using a global AIF [2]. In stroke cases, mul- leading to contrast-enhancement is commonly seen in clin-
tiple regional AIFs may diminish some of the dispersion ical practice, particularly with many brain tumors. In cases
errors [96] from vascular disease and using these along with of contrast agent leakage, underestimation of rCBV may
newer models based on vascular morphology and fluid occur because GBCA leakage can diminish the magnitude
dynamics [97, 98] holds promise to combat dispersion of the susceptibility contrast signal intensity loss in regions
effects. where T1 effects are prominent (Fig. 2.5). At the same
time, it is possible to overestimate rCBV in the face of
prominent T2/T2* effects because this will result in greater
Absolute Quantification signal decrease and undershooting of the baseline signal
intensity. The amount of under- or overestimation of rCBV
Absolute quantification of DSC-MRI is difficult due to sev- is contingent upon contrast agent kinetics, brain tissue
eral factors such as uncertainties relating to hematocrit, brain microstructure, pulse sequence parameters, and preload
proton density constants, contrast agent relaxivity, contrast GBCA dose [2].
24 M. S. Shiroishi et al.

Fig. 2.4 Example of spin and

gradient echo (SAGE)-based
DSC-MRI maps in a
glioblastoma patient showing
post-contrast T1-weighted
and fluid-attenuated inversion
recovery (FLAIR). As would
be expected, the tumor CBV,
CBF, MTT, and VSI values
are higher than those found in
contralateral normal-­
appearing white matter
(NAWM). Also note the
differences between GRE and
SE maps within the tumor,
particularly for CBF and
MTT. The Ktrans and CTH
maps also exhibit regional
heterogeneity within the
tumor. Such differences
highlight the unique and
complementary nature of
multi-echo SAGE
hemodynamic and vascular
sensitivity. For clarity, relative
parameter maps are shown
using the illustrated colorbar.
Reprinted with permission
from [2]
2 Physical Principles of Dynamic Contrast-Enhanced and Dynamic Susceptibility Contrast MRI 25

a b

Signal Intensity

Signal Intensity
Time Time

Fig. 2.5 (a) Schematic of signal intensity-time curve with BBB leak- Schematic of signal intensity-time curve with BBB leakage and domi-
age and dominant T1 leakage effect. T1-related signal enhancement nant T2/T2* effects. T2/T2* effects result in more signal decrease and
results in a less signal decrease and subsequent overshooting of the subsequent undershooting of the baseline (straight line). This will lead
baseline (straight line). This will lead to underestimation of rCBV. (b) to overestimation of rCBV. Reprinted with permission from [50]

The decreased susceptibility differences between the Superparamagnetic contrast agents such as iron oxide
intra- and extravascular compartments due to GBCA leakage nanoparticles are a newer type of contrast agent that may be
result in temporally variant decreases in GBCA T2* relaxiv- advantageous compared to GBCAs given their lack of
ity [25, 103]. More T2* signal decrease can result from extravasation, more prominent T2 and T2* relaxivity, and
mesoscopic magnetic field gradients induced by compart- recent concerns about potential long-term effects of gado-
mentalization of GBCA around cells (Fig. 2.6). These linium tissue deposition [114]. Though there are no current
changes may be influenced by cellular features such as such blood pool agents approved for DSC-MRI, ferumoxytol
shape, size, density, polydispersity, and atypia [104]. The can be used off-label for DSC-MRI [115]. Ferumoxytol is a
potential interaction between T1 and T2/T2* effects in the macromolecular, carbohydrate-coated iron oxide particle
same lesion further complicates interpretation of rCBV val- that has been sold under the name Feraheme as an iron
ues [105]. replacement for adult renal failure patients [116, 117], and
There has historically been various methodologies several studies have shown promise of DSC-MRI using this
employed to address leakage effects including low flip angle agent to distinguish pseudoprogression from tumor progres-
and dual TE acquisitions, preload GBCA dosing, and math- sion in brain tumor patients [118, 119].
ematical post-processing models [105–108]. Current recom-
mendations (see section “Standardization Efforts and
Variability of DSC-MRI” below) to correct for leakage  tandardization Efforts and Variability
S
effects in single-echo GRE-EPI sequences are to use 60° FA of DSC-MRI
acquisition with full-dose preload or 30° FA without preload,
both with full-dose bolus GBCA administration and applica- As with DCE-MRI, there has been a lack of standardized
tion of the Boxerman-Schmainda-Weisskoff (BSW) model-­ methodology for DSC-MRI [120]. Significant variation in
based leakage correction method [51, 70, 109–111]. The rCBV values can result from differences in image acquisi-
BSW model generates rCBV corrected for T1 and T2* leak- tion and post-processing methods, particularly with regard to
age effects by using linear fitting to calculate voxel-wise dif- dealing with leakage effects [121]. This lack of standardiza-
ferences in ΔR2∗ curves from non-enhancing regions and tion has made inclusion of DSC-MRI into clinical trials and
assumes unidirectional GBCA extravasation [105, 112]. routine practice rather difficult, and to help address this, the
Recent work in a rat glioma model suggests that dual-echo American Society of Functional Neuroradiology (ASFNR)
DSC-MRI acquisitions along with a combined biophysical published its recommended DSC-MRI protocol in 2015 cen-
and pharmacokinetic method can potentially eliminate the tered around 1/4–full-dose preload GBCA administration, an
need for preload GBCA dosing [113]. intermediate (60°) flip angle, field strength-dependent TE,
Another Random Scribd Document
with Unrelated Content
 CARTA 211
De Santa Teresa a su confesor Fray Jerónimo Gracián, llorando la
muerte del General de los Carmelitas Fray Juan Bautista Rubeo.
Fecha en Ávila a 15 de octubre de 1578.

Jesús.
Sea con vuestra paternidad el Espíritu Santo, mi padre[330]. Como
le veo quitado[331] de esas baraúndas, háseme quitado la pena de lo
demás, venga lo que viniere. Harto grande me la ha dado[332] las
nuevas, que me escriben de nuestro padre general. Ternísima estoy;
y el primer día llorar que llorarás[333], sin poder hacer otra cosa, y
con gran pena de los trabajos que le hemos dado, que cierto no los
merecía; y si hubiéramos ido a él, estuviera todo llano. Dios perdone
a quien siempre lo ha estorbado, que con vuestra paternidad yo me
aviniera, anque, en esto, poco me ha creído. El Señor lo trairá todo a
bien; mas siento lo que digo, y lo que vuestra paternidad ha
padecido; que cierto son tragos de la muerte lo que me escribió en
la carta primera, que dos he recibido después que habló al nuncio.
Sepa, mi padre, que yo me estaba deshaciendo, porque no daba
luego aquellos papeles, sino que debe ser aconsejado de quien le
duele poco lo que vuestra paternidad padece[334]. Huélgome, que
quedará bien experimentado, para llevar los negocios por el camino
que han de ir, y no agua arriba, como yo siempre decía: y a la
verdad ha habido cosas por donde lo impedían todo, y ansí no hay
que tratar de esto, porque ordena Dios cosas para que padezcan sus
siervos.
Ya quisiera escribir más largo, y han de llevar esta noche las
cartas, y casi lo es ya, que lo he sido[335] con el obispo de Osma[336],
para que trate con el presidente y con el padre Mariano lo que le
escribí, y dije enviase a vuestra paternidad. Ahora he estado con mi
hermano[337], y se le encomienda mucho.
 NOTAS
[305] Acordarse, construído como recordar con un dativo reflexivo y un
acusativo, es poco usado,
Y como Ovidio escribe en su epistolio,
que no me acuerdo el folio,
estas heridas del amor protervas
no se curan con hierbas.
Lope, Gatom. 2.

[306] Sobre pues, conjunción continuativa que encabeza las transiciones,

v. Bello. Gram. § 1267.

[307] Anticuado por entrambos. Esta cláusula juntábamonos entramos a

leer vidas de santos está sin duda trastocada, debiendo colocarse detrás
de gran amor y ellos a mí.

[308] Anque, forma vulgar por «aunque». Después hallaremos an por

«aún».

[309] Sino que en el sentido de pero. (V. Bello. Gram. § 1280.)

[310] Nótese a cada paso la ausencia de retoque; este complemento con

muchas lágrimas debiera ir inmediatamente después del verbo.

[311] Después de oraciones temporales, que puede usarse en vez de la

frase adverbial de tiempo luego que, después que; por ejemplo: «en
estando lejos de aquí, que me vea libre del peligro, no me meteré yo en
otra.» Si la oración temporal no lleva el verbo en gerundio ni infinitivo, sino
en forma personal, el que es un tanto pleonástico, pues pudiera
reemplazarse por la simple conjunción copulativa: «cuando esté lejos de
aquí, que (y) me vea libre...» Por este mismo giro se explican modismos
tales como estos: «jura que al volver que vuelva al Andalucía, se ha de
estar dos meses en Toledo»; «en llegando que llegue.»

[312] Este lo representa un adjetivo que no existe; Santa Teresa tomó en

su imaginación el substantivo de virtud por el adjetivo equivalente
virtuoso.

[313] Es muy común decir libros de caballería; ha de decirse caballerías

en plural, que este nombre se da a las hazañas llevadas a cabo por un
caballero. La afición a las novelas caballerescas fué predominante en
España por el espacio increíble de más de tres siglos. En el siglo XIV el
Canciller Pero López de Ayala, entre sus yerros más grandes, se lamentaba
de haber sido víctima de tan desatinada afición:
Plogome otrosí oir muchas vegadas
Libros de devaneos e mentiras probadas:
Amadis, Lanzalote e burlas asacadas,
En que perdí mi tiempo a muy malas jornadas.
(Rimado de Palacio, copla 162.)
A mediados del siglo XVI Santa Teresa se acusa de igual pecado, y a
principios del XVII era todavía tan desmedido el apego a tales novelas, que
Cervantes, para amenguarlo, ridiculizó en su Quijote los extravíos que tan
dañosa lectura causaba.

[314] Este los se refiere a los libros de caballerías que, aunque hace
mucho se nombraron, no dejan de estar presentes a la memoria en todo
este pasaje. Otra vez vemos aquí la sintaxis de la Santa obedecer más a la
viveza de la imaginación que a la lógica gramatical.

[315] El pronombre ella se refiere a la madre aunque no se la haya

nombrado inmediatamente antes. Otra vez cabe la observación de la nota
anterior.

[316] Nuevo descuido de la autora que pensaba haber escrito antes me

hizo enfriar, o cosa parecida.

[317] Créame y los verbos que siguen en singular debieran ir en plural,

pues la Autora se dirige a sus monjas, como adelante se ve.

[318] Santa Teresa trata generalmente a las religiosas de su merced en

tercera persona de plural; aquí las habla en segunda persona de plural. Es
común, en escritores más cuidados, este cambio de tratamiento. Fray Luis
de Granada dice a la Virgen: «alegrate con esta esperanza y cesen ya tus
gemidos... Bien veo, señora, que no basta nada desto para consolaros».
(B. Aut. esp., VIII, pág. 82 b).

[319] Esta especie de superlativo formado mediante el prefijo re que

refuerza el sentido del adjetivo simple, es muy propio del castellano
(refino, relimpio, remucho, remejor); muchos escritores lo desdeñan por
familiar.

[320] Ante los adverbios más y menos usaban nuestros clásicos las
formas apocopadas muy, tan, cuán («cuán más agradable»), en vez de las
formas plenas mucho, tanto, cuanto, que son hoy de rigor (V. Bello Gram.
§ 1023).

[321] Las leyes lógicas de la concordancia exigirían se hacen más nobles

y aparejados; la licencia hoy tolerable sería se hace aparejada.

[322] Serna era el mandadero que llevaba las cartas de don Lorenzo.

[323] Francisco se llamaba el hijo mayor de don Lorenzo. La Santa era

naturalmente directora de los negocios espirituales de todas las personas
de su familia. Lorenzo había prometido obediencia a su hermana, como
luego se verá.

[324] Este la representa al substantivo carta que la autora consideraba

embebido en el verbo escribiere. (Recuérdese lo dicho página 148 n. 312 y
pág. 149, 314 y 315, y véase 153, n. 326.)

[325] El sujeto de este verbo no es Francisco, como parece, sino don

Lorenzo.

[326] Este las se refiere a las monjas de la comunidad.

[327] An es contracción vulgar por aun. Comp. arriba anque.

[328] Sobra el que para hacer sentido.

[329] Guardar sin complemento, con el sentido de «guardar la

abstinencia».

[330] Vocativo con el posesivo antepuesto.

[331] Quitar tiene aquí el sentido anticuado de libertar, eximir, que

subsiste en la frase «libre y quito».

[332] Concordancia viciosa.

[333] Frase adverbial, como llora que llora o llora que llorarás, para
denotar la continuidad de la acción.

[334] Habla aquí de las persecuciones de que era objeto la reforma de la

Orden que entonces se llevaba a cabo. El entregar los papeles de la visita
al Presidente del Consejo de Castilla fué un paso poco acertado que dió
lugar a conflictos en los que Gracián quedó comprometido.
[335] El lo se refiere a larga en escribir; es decir: «que he sido larga en
escribir al Obispo». La autora pensaba haber puesto antes: «ya quisiera
ser más larga en escribir», en vez de «quisiera escribir más largo».

[336] El Obispo de Osma, don Alonso Vázquez, confesor de la Santa en

Toledo.

[337] Don Lorenzo de Cepeda.

 FRAY LUIS DE LEÓN
(1527-1591)

Los dos primeros libros de los Nombres de Cristo se imprimieron

en 1583; los tres completos, en 1585. La perfecta casada, en 1586.
Como se ha visto, la prosa castellana contaba ya en el último
tercio del siglo XVI con muy notables cultivadores.
Fray Luis de León consideraba, sin embargo, que el idioma no
había logrado aún el cultivo esmerado y profundo de que era digno.
Claro es que no podía satisfacerle, aunque lo admiraba, el estilo
humilde, sencillo y descuidado de Santa Teresa; pero ya es más
chocante que, hablando del poco cultivo de la lengua, no dedique ni
una alabanza, ni un recuerdo, a su predecesor, Fray Luis de
Granada; el estilo de éste era un estilo oratorio que sin duda, no
contentaba al maestro León, por no encajar dentro del ideal de
perfección artística que él perseguía[338]. Así que se consideró a sí
mismo, más que como innovador, como padre de la prosa literaria, y
no le faltaba alguna razón.
El lenguaje de Fray Luis de Granada tenía solemnidad, elevación
y valentía; pero por estar aún el idioma poco diestro en la expresión
de razonamientos y pensamientos abstractos, no halla muchas veces
los recursos delicados de la construcción gramatical, y tiene algo de
desmañado y flojo. Por esto Fray Luis de León encontró que el
castellano encerraba tesoros aun no hallados de cadencia,
proporción, asiento y armonía.
Granada se esforzó en trabajar la frase, considerándola como un
silogismo, como un razonamiento o un apóstrofe; León le dedicó su
cuidado mirándola más especialmente como una obra de arte. Los
tratados del uno son como sermones puestos por escrito; los del
otro, como poesías redactadas en prosa[339]. El uno es más
elocuente, el otro más poeta; el uno es, en suma, orador, y el otro
escritor.
Fray Luis de León nos declara que su arte era en todo reflexivo y
meditado; arte de selección cuidadosa de palabras, y hasta de
letras; arte de cálculo y medida en la disposición de frases; arte en
todo diestro, esmerado y primoroso que nos ofrece la lengua
castellana ataviada con todos los elementos poéticos y musicales de
que es capaz, y levantada a la altura de las lenguas clásicas.
Él mismo declara también que su empeño principal fué poner en
el habla del vulgo número, abundancia, entonación y armonía. Sin
embargo, a veces usa períodos defectuosos, y esto principalmente
por construirlos tan largos que casi se rompe el enlace de su
comienzo con su remate[340]. Además, las conjunciones porque y
pues aparecen encabezando multitud de frases, con el pueril objeto
de encadenarlas materialmente a la que antecede, cuando de no
ligarlas de otra manera bastaría que esta trabazón corriera
solamente a cargo del pensamiento. En fin, pocas veces cae en la
tentación de buscar la falsa elegancia, puesta en moda ya desde el
siglo XV, de remitir afectadamente el verbo al fin de la proposición
(verbi gracia: «Con el calor del día y del sueño encendidos
demasiadamente y dañados», pág. 175).

NOMBRES DE CRISTO

INTRODUCCIÓN AL LIBRO III

Declara Fray Luis en qué procuró mejorar el lenguaje de sus escritos

sobre el ordinario y familiar.

Mas a los que dicen que no leen aquestos mis libros por estar en
romance[341] y que en latín los leyeran, se les responde que les debe
poco su lengua, pues por ella aborrecen lo que, si estuviera en otra,
tuvieran por bueno. Y no sé yo de dónde les nace el estar con ella
tan mal; que ni ella lo merece, ni ellos saben tanto de la latina que
no sepan más de la suya, por poco que della sepan, como de hecho
saben della poquísimo muchos. Y destos son los que dicen que no
hablo en romance, porque no hablo desatadamente y sin orden, y
porque pongo en las palabras concierto y las escojo y les doy su
lugar; porque piensan que hablar romance es hablar como se habla
en el vulgo, y no conocen que el bien hablar no es común, sino
negocio de particular juicio[342], ansí en lo que se dice, como en la
manera como se dice; y negocio que de las palabras que todas
hablan elige las que convienen y mira el sonido dellas, y aun cuenta
a veces las letras, y las pesa y las mide y las compone, para que, no
solamente digan con claridad lo que se pretende decir, sino también
con armonía y dulzura. Y si dicen que no es estilo para los humildes
y simples, entiendan que, así como los simples tienen su gusto, así
los sabios y los graves y los naturalmente compuestos no se aplican
bien a lo que se escribe mal y sin orden; y confiesen que debemos
tener cuenta con ellos, y señaladamente en las escrituras que son
para ellos solos, como aquesta lo es.
Y si acaso dijeren que es novedad, yo confieso que es nuevo, y
camino no usado por los que escriben en esta lengua, poner en ella
número, levantándola del decaimiento ordinario. El cual camino
quise yo abrir[343], no por la presunción que tengo de mí, que sé bien
la pequeñez de mis fuerzas, sino para que los que las tienen se
animen a tratar de aquí adelante su lengua como los sabios y
elocuentes pasados, cuyas obras por tantos siglos viven, trataron las
suyas, y para que la igualen, en esta parte que le falta, con las
lenguas mejores, a las cuales, según mi juicio, vence ella en otras
muchas virtudes.
 LIBRO PRIMERO
Dirigiéndose al Obispo de Córdoba, don Pedro Portocarrero,
introduce Fray Luis los personajes que figurarán en el diálogo de
la obra, y supone que son tres amigos suyos, de su misma Orden
de San Agustín.

Era por el mes de Junio, a las vueltas[344] de la fiesta de San

Juan, al tiempo que en Salamanca comienzan a cesar los estudios,
cuando Marcelo, el uno de los que digo (que así le quiero llamar con
nombre fingido, por ciertos respetos que tengo, y lo mismo haré a
los demás), después de una carrera tan larga, como es la de un año
en la vida que allí se vive[345], se retiró, como a puerto sabroso, a la
soledad de una granja que, como vuestra merced sabe, tiene mi
monasterio en la ribera de Tormes[346]; y fuéronse con él, por hacerle
compañía, y por el mismo respeto, los otros dos. Adonde habiendo
estado algunos días, aconteció que una mañana, que era la del día
dedicado al apóstol San Pedro, después de haber dado al culto
divino[347] lo que se le debía, todos tres juntos se salieron de la casa
a la huerta que se hace[348] delante della. Es la huerta grande, y
estaba entonces bien poblada de árboles, aunque puestos sin orden;
mas eso mismo hacía deleite en la vista, y sobre todo, la hora y la
sazón.
Pues entrados en ella, primero, y por un espacio pequeño, se
anduvieron paseando y gozando del frescor, y después se sentaron
juntos a la sombra de unas parras y junto a la corriente de una
pequeña fuente, en ciertos asientos. Nace la fuente de la cuesta que
tiene la casa a las espaldas, y entraba en la huerta por aquella
parte, y corriendo y estropezando, parecía reírse. Tenían también
delante de los ojos y cerca dellos una alta y hermosa alameda. Y
más adelante, y no muy lejos, se veía el río Tormes, que aun en
aquel tiempo, hinchiendo bien sus riberas, iba torciendo el paso por
aquella vega. El día era sosegado y purísimo, y la hora muy fresca.
Así que, asentándose y callando por un pequeño tiempo, después de
sentados, Sabino (que así me place llamar al que de los tres era el
más mozo), mirando hacia Marcelo y sonriéndose, comenzó a decir
así:
«Algunos hay a quien la vista del campo los enmudece[349], y
debe ser condición de espíritus de entendimiento profundo; mas yo,
como los pájaros, en viendo lo verde, deseo o cantar o hablar.»
—«Bien entiendo por qué lo decís—respondió al punto Marcelo—,
y no es alteza de entendimiento, como dais a entender por
lisonjearme o por consolarme, sino cualidad de edad y humores
diferentes que nos predominan y se despiertan con esta vista, en
vos de sangre, y en mí de melancolía[350]. Mas sepamos—dice—de
Juliano[351] (que éste era el nombre del tercero) si es pájaro también
o si es de otro metal.»
—«No soy siempre de uno mismo—respondió Juliano—, aunque
agora al humor de Sabino me inclino algo más. Y pues él no puede
agora razonar consigo mismo mirando la belleza del campo y la
grandeza del cielo, bien será que nos diga su gusto acerca de lo que
podremos hablar.»
Entonces Sabino, sacando del seno un papel escrito y no muy
grande: «Aquí, dice, está mi deseo y mi esperanza.»
Marcelo, que reconoció luego el papel, porque estaba escrito de
su mano, dijo, vuelto a Sabino y riéndose: «No os atormentará
mucho el deseo a lo menos, Sabino, pues tan en la mano tenéis la
esperanza; ni aun deben ser ni lo uno ni lo otro muy ricos, pues se
encierran en tan pequeño papel.»
—«Si fueren pobres—dijo Sabino—, menos causa tendréis para
no satisfacerme en una cosa tan pobre.»
—«¿En qué manera—respondió Marcelo—, o qué parte soy yo
para satisfacer a vuestro deseo, o qué deseo es el que decís?»
 Entonces Sabino, desplegando el papel, leyó el título, que decía:
De los nombres de Cristo; y no leyó más, y dijo luego: «Por cierto
caso hallé hoy este papel, que es de Marcelo, adonde, como parece,
tiene apuntados algunos de los nombres con que Cristo es llamado
en la Sagrada Escritura, y los lugares de ella adonde es llamado así.
Y como le vi, me puso codicia de oirle algo sobre aqueste
argumento, y por eso dije que mi deseo estaba en este papel; y está
en él mi esperanza también, porque, como parece dél, éste es
argumento en que Marcelo ha puesto su estudio y cuidado, y
argumento que le debe tener en la lengua; y así, no podrá decirnos
agora lo que suele decir cuando se excusa, si le obligamos a hablar,
que le tomamos desapercibido. Por manera que, pues le falta esta
excusa, y el tiempo es nuestro, y el día santo, y la sazón tan a
propósito de pláticas semejantes, no nos será dificultoso el rendir a
Marcelo, si vos, Juliano, me favorecéis.»

LIBRO II, CAPÍTULO III

Marcelo explicando a sus amigos por qué el nombre de Príncipe de
Paz es aplicado a Cristo, declara qué cosa es paz.

Calló Marcelo un poco, luego que dijo esto..., y descansando, y

como recogiéndose[352] todo en sí mismo por un espacio pequeño,
alzó después los ojos al cielo, que ya estaba sembrado de estrellas,
y teniéndolos en ellas como enclavados, comenzó a decir así:
«Cuando[353] la razón no lo demostrara, ni por otro camino se
pudiera entender cuán amable cosa sea[354] la paz, esta vista
hermosa del cielo que se nos descubre agora, y el concierto que
tienen entre sí aquestos resplandores que lucen en él, nos dan
suficiente testimonio. Porque, ¿qué otra cosa es, sino paz, o
ciertamente una imagen perfecta de paz, esto que agora vemos en
el cielo y que con tanto deleite se nos viene[355] a los ojos? Que[356] si
la paz es, como San Agustín breve y verdaderamente concluye, una
orden sosegada o un tener sosiego y firmeza en lo que pide el buen
orden, eso mismo es lo que nos descubre agora esta imagen.
Adonde el ejército de las estrellas, puesto como en ordenanza y
como concertado por sus hileras[357], luce hermosísimo; y adonde
cada una dellas inviolablemente guarda su puesto; adonde no
usurpa ninguna el lugar de su vecina ni la turba en su oficio, ni
menos, olvidada del suyo, rompe jamás la ley eterna y santa que le
puso la Providencia; antes, como hermanadas todas y como
mirándose entre sí, y comunicando sus luces las mayores con las
menores, se hacen muestra de amor; y como en cierta manera[358]
se reverencian unas a otras, y todas juntas templan a veces sus
rayos y sus virtudes, reduciéndolas a una pacífica unidad de virtud,
de partes y aspectos diferentes compuesta, universal y poderosa
sobre toda manera[359].
»Y si así se puede decir, no sólo son un dechado de paz clarísimo
y bello, sino un pregón y un loor que con voces manifiestas y
encarecidas nos notifica cuán excelentes bienes son los que la paz
en sí contiene y los que hace en todas las cosas. La cual voz y
pregón sin ruido se lanza en nuestras almas, y de lo que en ellas
lanzada hace[360], se ve y entiende bien la eficacia suya y lo mucho
que las persuade. Porque luego, como convencidas de cuanto les es
útil y hermosa la paz, se comienzan ellas a pacificar en sí mismas y a
poner a cada[361] una de sus partes en orden. Porque si estamos
atentos a lo secreto que en nosotros pasa, veremos que este
concierto y orden de las estrellas, mirándolo, pone en nuestras
almas sosiego, y veremos que con sólo tener los ojos enclavados en
él con atención, sin sentir en qué manera, los deseos nuestros y las
afecciones turbadas que confusamente movían ruido en nuestros
pechos de día, se van quietando poco a poco, y como
adormeciéndose, se reposan, tomando cada una su asiento, y
reduciéndose a su lugar propio, se ponen sin sentir en sujeción y
concierto.
»Y veremos que, así como ellas se humillan y callan, así lo
principal y lo que es señor en el alma, que es la razón, se levanta y
recobra su derecho y su fuerza, y como alentada con esta vista
celestial y hermosa, concibe pensamientos altos y dignos de sí, y
como en una cierta manera se recuerda[362] de su primer origen, y al
fin pone todo lo que es vil y bajo en su parte, y huella sobre ello[363].
Y así puesta ella en su trono como emperatriz, y reducidas a sus
lugares todas las de más partes del alma, queda todo el hombre
ordenado y pacífico.
«Mas ¿qué digo de nosotros que tenemos razón? Esto insensible
y aquesto rudo del mundo, los elementos y la tierra y el aire y los
brutos se ponen todos en orden y se quietan luego que poniéndose
el sol, se les representa aqueste ejército resplandeciente. ¿No veis el
silencio que tienen agora todas las cosas, y cómo parece que
mirándose en este espejo bellísimo, se componen todas ellas y
hacen paz entre sí, vueltas a sus lugares y oficios, y contentas con
ellos?
»Es sin duda el bien de todas las cosas universalmente la paz; y
así, dondequiera que la ven, la aman. Y no sólo ella, mas la vista de
su imagen de ella las enamora y las enciende en codicia de
asemejársele, porque todo se inclina fácil y dulcemente a su bien. Y
aun si confesamos, como es justo confesar, la verdad, no solamente
la paz es amada generalmente de todos, mas sola ella es amada y
seguida y procurada por todos. Porque cuanto se obra en esta vida
por los que vivimos en ella, y cuanto se desea y afana, es por
conseguir este bien de la paz, y este es el blanco adonde enderezan
su intento y el bien a que aspiran todas las cosas. Porque si navega
el mercader y si corre los mares, es por tener paz con su codicia,
que le solicita y guerrea. Y el labrador en el sudor de su cara y
rompiendo la tierra busca paz, alejando de sí cuanto puede al
enemigo duro de la pobreza. Y por la misma manera, el que sigue el
deleite y el que anhela la honra y el que brama por la venganza, y,
finalmente, todos y todas las cosas buscan la paz en cada una de
sus pretensiones. Porque, o siguen algún bien que les falta, o huyen
algún mal que los enoja.»

LA PERFECTA CASADA

LIBRO VII

Comentando el versículo de los Proverbios, XXXI, 15: «madrugó y

repartió a sus gañanes las raciones», hace Fray Luis una
primorosa descripción del alba y encarece las delicias del
madrugar.

El madrugar es tan saludable, que la razón sola de la salud,

aunque no despertara el cuidado y obligación de la casa, había de
levantar de la cama en amanesciendo a las casadas. Y guarda en
esto Dios, como en todo lo demás, la dulzura y suavidad de su sabio
gobierno, en que aquello a que nos obliga es lo mismo que más
conviene a nuestra naturaleza y en que recibe por su servicio lo que
es nuestro provecho[364]. Así que, no sólo la casa, sino también la
salud, pide a la buena mujer que madrugue. Porque cierto es que es
nuestro cuerpo del metal de los otros cuerpos, y que la orden que
guarda la naturaleza para el bien y conservación de los demás, esa
misma es la que conserva y da salud a los hombres.
Pues ¿quién no ve que a aquella hora despierta el mundo todo
junto, y que la luz nueva saliendo, abre los ojos de los animales
todos, y que si fuese entonces dañoso dejar el sueño, la naturaleza
(que en todas las cosas generalmente, y en cada una por sí, esquiva
y huye el daño, y sigue y apetece el provecho, o que, para decir la
verdad, es ella eso mismo que a cada una de las cosas conviene y es
provechoso), no rompiera tan presto el velo de las tinieblas que nos
adormecen, ni sacara por el oriente los claros rayos del sol, o si los
sacara, no les diera tanta fuerza para nos despertar?[365]. Porque si
no despertase naturalmente la luz, no le cerrarían las ventanas tan
diligentemente los que abrazan el sueño. Por manera que la
naturaleza, pues nos envía la luz, quiere, sin duda, que nos
despierte. Y pues ella nos despierta, a nuestra salud conviene que
despertemos.
Y no contradice a esto el uso de las personas que ahora el
mundo llama señores, cuyo principal cuidado es vivir para el
descanso y regalo del cuerpo, las cuales guardan la cama hasta las
doce del día[366]. Ante esta verdad, que se toca con las manos,
condena aquel vicio, del cual, ya por nuestros pecados o por sus
pecados de ellos mismos[367], hacen honra y estado[368], y ponen
parte de su grandeza en no guardar ni aun en esto el concierto que
Dios les pone. Castigaba bien una persona, que yo conocí, esta
torpeza, y nombrábala con su merescido vocablo. Y aunque es tan
vil como lo es el hecho, daráme vuestra merced[369] licencia para que
lo ponga aquí, porque es palabra que cuadra. Así que, cuando le
decía alguno que era estado en los señores este dormir, solía él
responder que se erraba la letra[370], y que por decir establo decían
estado. Y ello a la verdad es así, que aquel desconcierto de vida
tiene principio y nasce de otro mayor desconcierto, que está en el
alma y es causa él también y principio de muchos otros
desconciertos torpes y feos. Porque la sangre y los demás humores
del cuerpo, con el calor del día y del sueño, encendidos
demasiadamente y dañados, no solamente corrompen la salud, mas
también aficionan e inficionan el corazón feamente. Y es cosa digna
de admiración que, siendo estos señores en todo lo demás grandes
seguidores, o por mejor decir, grandes esclavos de su deleite, en
esto sólo se olvidan dél, y pierden por un vicioso dormir lo más
deleitoso de la vida, que es la mañana.
Porque entonces la luz, como viene después de las tinieblas y se
halla como después de haber sido perdida, parece ser otra y hiere el
corazón del hombre con una nueva alegría, y la vista del cielo
entonces, y el colorear de las nubes y el descubrirse el aurora (que
no sin causa los poetas la coronan de rosas)[371], y el aparecer la
hermosura del sol, es una cosa bellísima. Pues el cantar de las aves,
¿qué duda hay sino que suena entonces más dulcemente? y las
flores y las yerbas y el campo, todo despide de sí un tesoro de olor.
Y como cuando entra el rey de nuevo en alguna ciudad se adereza y
hermosea toda ella, y los ciudadanos hacen entonces plaza[372] y
como alarde de sus mejores riquezas; así los animales y la tierra y el
aire, y todos los elementos, a la venida del sol se alegran, y como
para recibirle, se hermosean y mejoran y ponen en público cada uno
sus bienes. Y como los curiosos suelen poner cuidado y trabajo por
ver semejantes recibimientos, así los hombres concertados y
cuerdos, aun por sólo el gusto, no han de perder esta fiesta que
hace toda la naturaleza al sol por las mañanas; porque no es gusto
de un solo sentido, sino general contentamiento de todos, porque la
vista se deleita con el nascer de la luz y con la figura[373] del aire y
con el variar de las nubes; a los oídos las aves hacen agradable
armonía; para el oler, el olor que en aquella sazón el campo y las
yerbas despiden de sí es olor suavísimo, pues el fresco del aire de
entonces templa con grande deleite el humor calentado con el
sueño, y cría salud y lava las tristezas del corazón, y no sé en qué
manera le despierta a pensamientos divinos antes que se ahogue en
los negocios del día.
Pero, si puede tanto con estos hijos de tinieblas el amor dellas,
que aun del día hacen noche, y pierden el fruto de la luz con el
sueño, y ni el deleite, ni la salud, ni la necesidad y provecho que
dicho habemos, son poderosos para los hacer levantar, vuestra
merced que es hija de luz, levántese con ella, y abra la claridad de
sus ojos cuando descubriere sus rayos el sol, y con pecho puro
levante sus manos limpias al Dador de la luz, ofresciéndole con
santas y agradescidas palabras su corazón, y después de hecho esto,
y de haber gozado del gusto del nuevo día, vuelta a las cosas de su
casa, entienda en su oficio.

NOTAS
[338] Véase la nota 343 de la pág. 161.

[339] Algunos de sus párrafos tienen el mismo asunto que sus versos, no
sabiéndose si son su esbozo y plan o su comentario y explicación. (Véase
pág. 169, nota 359, y pág. 170, nota 363.)

[340] Véase, por ejemplo, la larga interrogación de la pág. 173.

[341] Se censuró a Fray Luis por haber escrito en castellano los dos
primeros libros de los Nombres de Cristo, impresos en 1583; pues, aunque
ya habían escrito el P. Avila y el P. Granada, muchos seguían creyendo que
un teólogo no debía emplear para sus obras sino el latín. Fray Luis
contestó reimprimiendo los Nombres de Cristo, en 1585, adicionados con
un tercer libro a cuya introducción pertenece el presente extracto.

[342] Es decir, que no es cosa común a todos los que hablan una lengua,
sino que exige particular disposición y estudio. Es antigua en España la
creencia de que la lengua propia ni merece ni requiere atención y trabajo;
Juan de Valdés se queja de los que con tanta negligencia y tan inmerecido
desdén la tratan, y Ambrosio de Morales, en 1546, decía: «siempre ha
quedado nuestra lengua en la miseria y con la pobreza que antes tenía...
que todo nace del gran menosprecio en que nuestros mismos naturales
tienen nuestra lengua, por lo cual ni se aficionan a ella, ni se aplican a
ayudarla». (Introducción al Diálogo de la dignidad del hombre, del M.
Hernán Pérez de Oliva, tío de Morales.)

[343] Fray Luis, al principio de esta introducción, habla poco menos que
como si él fuera el primero en aplicar el castellano a asuntos serios,
quejándose «de lo mal que usamos de nuestra lengua no la empleando
sino en cosas sin ser». No es admisible que desconociera los autores
citados en la pág. 125, y por fuerza habría leído las obras místicas del
Beato Juan de Ávila y del Venerable Granada, que andaban ya impresas;
sin embargo, a juzgar por las palabras que ahora emplea, parece que no le
satisfacían mucho y no las tomaba en consideración.

[344] A vueltas de significa ‘alrededor de, cerca de’; así fijando después
el día en que esto sucedía, dícese que era el de San Pedro, que es en 29
de Junio, cinco días después de San Juan. En esta frase el artículo se usa
rarísima vez: a las vueltas.

[345] Cuando el acusativo es de igual raíz que el verbo, exige algún

complemento que le especifique, pues de lo contrario sería un acusativo
del todo inútil, v. gr.: vivir una vida fatigosa (véase Bello, Gram. § 796);
aquí se sobreentiende con la vida (tan fatigosa) que allí se vive.

[346] Los nombres de ríos sin artículo, v. pág. 86, n. 161. Los agustinos
calzados, que llegaron a Salamanca por los años 1330, fueron los
fundadores de este convento. Hoy no existe el edificio antiguo, pues fué
bárbaramente destruído por el ejército francés en 1812, y aunque
reedificado, se demolió más tarde, ocupando hoy su solar la nueva calle
llamada de Oliva.—Este monasterio tenía, para descanso y recreo de los
frailes, una granja, llamada la Flecha, a legua y media de distancia, río
arriba, a la vera del camino de Salamanca a Madrid. (V. M. Villar y Macías,
Hist. de Salamanca, I, 453, etc.) La apacible descripción que hace Fray
Luis de este paisaje concuerda en todo con la realidad; tal como él lo
pinta, se reconocen hoy la casa de los frailes, las cuestas que empiezan a
sus espaldas y que si hacia Aldealengua se van insensiblemente
suavizando y disminuyendo, prolónganse larguísimo espacio
eslabonándose hacia Salamanca; todavía existe la desordenada arboleda
que tanto deleitaba la vista del poeta, y la risueña fuente que baja desde
la cuesta al huerto,
y como codiciosa
de ver y acrecentar su hermosura,
hasta llegar, corriendo se apresura.
En fin, el huerto mismo existe, que tanta inspiración guardaba para el
autor de la oda a la Vida retirada y que se llama, como queda dicho,
huerta de la Flecha.

[347] Destinada al culto está desde antiguo una capilla cerca de la

huerta, frente a la aceña de la Flecha y contigua a la casa del molinero.
[348] Hacerse era muy usado con nombres de lugar en el mismo sentido
que ‘extenderse, hallarse’, o sea ‘estar situado’.

[349] Los dice la edición de Salamanca 1585. Es el acusativo que debe

ponerse con propiedad gramatical; pero disuena algo a causa del uso
generalísimo del dativo le por el acusativo, cuando se trata de personas.

[350] Humor de sangre y de melancolía significa temperamento

sanguíneo y melancólico o bilioso.

[351] Sepamos de Juliano si es pájaro, en vez de sepamos si Juliano es

pájaro, es un caso de atracción del sujeto de la proposición dependiente
que se construye con el verbo principal; como en griego y en latín: rem
vides quomodo se habeat (v. Diez, Gr. III, 360.)

[352] Nótese el uso que tiene el adverbio como; como recogiéndose no

afirma que se recogiera sino que todo su aspecto y semejanza era como la
del que se recoge; como enclavados, semejando enclavados; como viene a
ser en ambos ejemplos un simple afijo o partícula prepositiva para denotar
mera semejanza con la voz que le sigue, sentido que se ve más claro si el
como se refiere a un substantivo: «encontró Don Quijote con dos como
clérigos», «unos como joyeles» (v. Bello, Gramática, § 1234 y 1236).

[353] Cuando tiene muchas veces el valor de la frase adverbial aun

cuando.

[354] En las interrogaciones indirectas la proposición secundaria puede

llevar su verbo en indicativo (como hoy es lo ordinario) o en subjuntivo;
aquí se diría hoy más bien: «cuán amable cosa es la paz». En los siglos XVI
y XVII era más común el subjuntivo, «dícese qué cosa sea la paz, lo que
valga la paz».

[355] Venirse a los ojos equivale a ‘saltar a la vista’ o ‘presentarse’.

[356] Que, conjunción causal, abreviada de porque.

[357] Respecto al como repetidas veces usado aquí para denotar no el

modo, sino la semejanza con ese modo, véase la nota 352, de la pág. 167:
como mirándose, semejando que se miran. Concertado por sus hileras se
diría simplemente hoy: «concertado por hileras» (o sea distribuído en
hileras), sin el posesivo; éste indica que el concierto les es a las estrellas
propio y natural. Es modismo antiguo; Don Alfonso el Sabio dice «fabla el
Arzobispo por su latín», es decir: en el latín que usaba siempre al escribir.
[358] Hoy este como que denota semejanza no se suele usar antepuesto
a verbos y proposiciones enteras, sino después de verbos que denotan una
apreciación o figuración; es decir, seguido de un que enunciativo: «se me
figuraba como que querían acercarse aquellos hombres», «hace como
que no quiere». «Como en cierta manera se reverencian», sería hoy:
«parece como que se reverencian»; al fin de este trozo se repite este
mismo giro: como en una cierta manera recuerda = ‘parece como que
recuerda’.

[359] Esta admirable descripción recuerda y amplía algunos versos de la

Oda XII del mismo autor, «Noche serena»:
Quién mira el gran concierto
de aquestos resplandores eternales,
su movimiento cierto,
sus pasos desiguales,
y en proporción concorde tan iguales...

[360] Lanzar, echar pregón o voz se emplean por los simples ‘pregonar’ o
‘vocear’. Compárese la concordancia voz y pregón lanzada con la que
hallamos en la Introducción al Símbolo de la fe (pág. 142) y en el Quijote
(comienzo del extracto de la parte II, capítulo 23).

[361] A cada se lee en la edición de Salamanca, 1585. Antes se admitían

más acusativos con preposición; hoy apenas se le pone a sino cuando el
acusativo es nombre de persona determinada, personificación, animal o
nombre propio de lugar, así que se diría «a poner cada una de sus partes».
También se diría con más rigor: «comienzan ellas a pacificarse y a poner
sus partes en orden», pues la acción reflexiva no se refiere para nada a
poner y sí sólo a pacificar, por lo cual no debe agregarse el pronombre
reflexivo a comienzan, ya que este verbo rige lo mismo a poner que a
pacificar.

[362] Para el giro como en cierta manera, véase la nota 358, pág. 168.
Acordarse y recordarse tenían, como se ve aquí, una misma construcción y
régimen (cfr. p. 145, n. 305). Hoy se diferencia mucho, pues se dice
acordar-se de una cosa y recordar una cosa.

[363] El alma contemplando la hermosura de la noche estrellada se

acuerda de su primer origen que es celestial, se siente como desterrada en
este mundo y ve con claridad las alturas del otro. Igual pensamiento
expuso en verso el maestro León, y casi con iguales palabras que aquí,
salvo que no es el espectáculo de la noche serena el que arroba el alma,
sino la sublime música del ciego Francisco Salinas:
 A cuyo son divino
mi alma, que en olvido está sumida,
torna a cobrar el tino
y memoria perdida
de su origen primera esclarecida.
Y como se conoce,
en suerte y pensamientos se mejora,
el oro desconoce
que el vulgo ciego adora,
la belleza caduca engañadora...

[364] Esto es, «en que agradece como un servicio lo que debemos hacer
por nuestro provecho».

[365] Hoy los pronombres personales átonos nunca se anteponen al

infinitivo, sino que se le posponen enclíticos. (V. Bello Gram. § 915). Fray
Luis de Granada dice «que nadie sea osado a la despertar». (Guía de pec.
I. 16. § 1 B. AA. EE. VI, 61 a.) Sólo como provincialismo se conserva la
costumbre arcaica; en Asturias, por ejemplo, se puede decir: «hay que lo
dejar», «tengo que os contar».

[366] Este es antiguo defecto español atestiguado por algunos

extranjeros; el barón alemán Conrado de Bemelberg, que para
perfeccionarse en el castellano viajó por España ocho años después de
muerto Fray Luis, escribe en una carta, fecha en agosto de 1599, dando
cuenta a su padre de lo que le parecía nuestra tierra: «quien en España
quiere negociar, más que ordinaria paciencia ha de tener, pues a mediodía
tienen costumbre de levantarse, y después de levantados ir a la misa,
acabada la cual se meten a comer, y después de la comida, o a jugar o a
dormir o pasearse a caballo por las calles».

[367] En sus pecados de ellos no es de ellos un inútil pleonasmo, sino

que está exigido por la vaguedad del su, que no determina si el poseedor
es masculino o femenino, ni singular o plural. Hoy esta doble indicación del
posesivo no se conserva sino cuando el poseedor es usted: «su padre de
usted», «su casa de usted».

[368] Nótese la frase, no registrada en los Diccionarios: hacer honra y

estado de una cosa, ‘fundar en ella su condición y su dignidad’.

[369] Vuestra merced se dirige a Doña María Varela Osorio, a la cual

dedicó su obra Fray Luis de León.
[370] Errar la letra es frase figurada; tómase en sentido propio
«equivocarse en la escritura o lectura», cuando se trata de algún
documento escrito, sobre cuya interpretación se discute. El uso de esta
expresión, u otras análogas, era muy corriente. En la Celestina (auto IX) se
dice, hablando de las veces que se debe beber: «Madre, pues tres veces
dicen que es lo bueno y honesto todos los que escribieron.—Hijos, estará
corrupta la letra: por trece, tres.» (Véase Rev. de Filología Española, IV,
50).

[371] Homero calificó a la Aurora de dedos de rosa y según él todos los

poetas clásicos; Ovidio llámala rosea dea (Ars. am. III. 84). Claro es que
en el Renacimiento esta denominación era un lugar común. Cervantes la
llamó rosada aurora (Quijote I. 2).

[372] Hacer plaza no está registrado en los diccionarios con el sentido

que aquí tiene de ‘hacer ostentación’. Sólo se le apunta el significado de
‘sacar a la plaza o publicar una cosa’.

[373] Figura dice la edición de Salamanca 1586, pero debe ser errata.
 EL P. JUAN DE MARIANA
(1536-1623)

Su Historia de España latina salió a luz por primera vez en Toledo

en 1592; en la misma ciudad se publicó la primera edición
romanceada en 1601.
La historiografía contaba ya en España con diestros
investigadores, que habían rectificado multitud de errores de la
historia tradicional, mediante el estudio crítico de crónicas, diplomas,
inscripciones, etc.; tales eran Garibay, Ambrosio de Morales, Zurita.
Mariana no se sentía inclinado a estas tareas, pues las suyas
habituales eran las del teólogo y moralista; sólo como ocupación
accesoria se dedicó a componer la Historia de España. Así que no se
propuso continuar los estudios especiales en averiguación de la
verdad, sino que, contentándose con lo hecho por otros, como en
sus obras echaba de menos el arte de la narración, no aspiraba sino
a vulgarizar lo estudiado por otros: mi intento no fué hacer historia,
sino poner en orden y estilo lo que otros habían recogido. Su
principal preocupación fué, pues, la narración agradable; escoge en
las diversas fuentes que maneja la versión de los hechos que
buenamente le parece más verdadera, y luego la expone sin reparo
crítico alguno; sucediendo más de una vez que la hermosura de un
relato fabuloso le atrae y le obliga a acogerlo sin expresar la menor
duda, pues lo que él pretendía era hacer, más que una historia
averiguada, una historia literaria y nacional, de la cual nada bello y
nada heroico debía ser excluído. Ciertamente que consiguió tal
propósito; su obra es hasta ahora el más digno monumento en
honor de la historia y tradiciones españolas, como lo es Tito Livio de
las romanas.
En el estilo de esta obra se ven claramente influencias, tanto de
la índole personal del autor, como de sus lecturas habituales. La
entereza de carácter y la austeridad de pensamiento de Mariana se
reflejan en su narración histórica, a veces seca, pero que sabe
revestirse siempre de un aire de autoridad y decoro que, como dice
Capmany, «apenas distingue uno después si son las cosas o las
palabras las que aparecen grandes y majestuosas». Ni aun en las
arengas es declamador o retórico.
Las habituales tareas de teólogo, político y moralista a que se
consagró Mariana, hacen que su narración, no sólo esté llena de
máximas y aforismos, según la costumbre general de los
historiadores de la época, sino que se desvíe, más o menos
visiblemente, para obligarla a correr por el cauce de las ideas
filosóficas y sociales del autor.
Su cultura clásica le hace imitar a Tito Livio en la manera amplia
y tranquila de relatar, y a Tácito en las sentencias y reflexiones con
que moraliza constantemente el relato. Además, como Mariana había
escrito primero su obra en latín, de aquí que al romancearla
conservara algún dejo de construcción latina como el que
apuntamos en la nota de la página 193.
En fin: la obligada lectura de crónicas castellanas de los siglos XIV
y XV le encariñó con el lenguaje viejo, y de ellas se le pegaron
multitud de arcaísmos, como: aína ‘presto, luego’; al ‘otro’, asaz
‘bastante, harto’; ca ‘porque’, muy usado por Mariana, y algo
también por Fray Luis de Granada; dende ‘desde allí’, hobo ‘hubo’,
maguer ‘aunque’, suso ‘arriba’. Sin duda esto tenía por objeto
revestir así el lenguaje de un aspecto más venerable. Razón tenía
Saavedra Fajardo al decir en su República literaria que así como
otros se tiñen las barbas por parecer mozos, Mariana se las teñía por
hacerse viejo. Lo cierto es que con ser la Historia de España treinta
años posterior a la Guerra de Granada de Mendoza, representa un
lenguaje mucho más antiguo. Este no es defecto especial de
Mariana, quien sabe mantener en un límite prudente el arcaísmo; las
Crónicas ejercían tal atractivo sobre los que las leían, que los poetas
que sacaban de ellas romances o comedias, solían imitar su lenguaje
arcaico con mucha más exageración que a Mariana, pues llegaban a
escribir sus versos contrahaciendo la fabla antigua.
Además del arcaísmo prudentemente manejado, se observa en
Mariana alguna otra afectación; sobre todo un particular estudio
para huir del uso del gerundio, forma verbal de que tanto abusan las
malas narraciones; en su lugar, Mariana emplea con preferencia el
participio oracional. Fuera de esto, el estilo de Mariana se distingue
por una gran llaneza y naturalidad, y por una construcción ligera que
prefiere la nueva yuxtaposición de las cláusulas a englobarlas con
relación de dependencia[374].

H I S TO R I A D E E S PA Ñ A

LIBRO XVII, CAPÍTULO XIII

Muerte del Rey Don Pedro el Cruel, 22 ó 23 marzo, 1369. En el

capítulo anterior contó Mariana cómo Don Enrique, vuelto de
Francia, allegó en rededor suyo muchos partidarios; le recibieron
por Rey Burgos y otras ciudades, y cercó a Toledo que aún se
mantenía por Don Pedro.

El Rey Don Pedro, desamparado de los que le podían ayudar, y

sospechoso de los demás, lo que sólo le restaba, se resolvió de
aventurarse, encomendarse a sus manos y ponerlo todo en el trance
y riesgo de una batalla; sabía muy bien que los reinos se sustentan y
conservan más con la fama y reputación que con las fuerzas y
armas. Teníale con gran cuidado el peligro de la real ciudad de
Toledo; estaba aquejado y pensaba cómo mejor podría conservar su
reputación. Esto le confirmaba más en su propósito de ir en busca
de su enemigo y dalle[375] la batalla. Procuráronselo estorbar los de
Sevilla; decíanle que se destruía y se iba derecho a despeñar; que lo
mejor era tener sufrimiento, reforzar su ejército y esperar las gentes
que cada día vendrían de sus amigos y de los pueblos que tenían su
voz[376]. Esto que le aconsejaban era lo que en todas maneras
debiera seguir, si no le cegaran la grandeza de sus maldades y la
divina justicia, que estaba ya determinada de muy presto castigallas.
Estando en este aprieto, sucedióle otro desastre, y fué que Vitoria,
Salvatierra y Logroño, que eran de su obediencia, fatigadas de las
armas del Rey de Navarra[377], y por falta de socorro por estar Don
Pedro tan lejos, se entregaron al Navarro. Ayudó a esto Don
Tello[378], el cual, si estaba mal con Don Pedro, no era amigo de su
hermano Don Enrique, y así se estaba a la mira[379] en Vizcaya, sin
querer ayudar a ninguno de los dos.
Proseguíase en este comedio el cerco de Toledo. Y como quier
que aquella ciudad estuviese, como dijimos, dividida en aficiones,
algunos de los que favorecían a Don Enrique intentaron de
apoderalle[380] de una torre del muro de la ciudad que miraba al real,
que se dice la torre de los Abades. Como no le sucediese[381] esta
traza, procuraron dalle entrada en la ciudad por el puente de San
Martín[382], sobre lo cual los del un bando y del otro vinieron a las
manos, en que sucedieron algunas muertes de ciudadanos.
Sabidas estas revueltas por el Rey Don Pedro, dióse muy mayor
priesa a irla a socorrer, por no hallalla perdida cuando llegase. Para ir
con menor cuidado mandó recoger sus tesoros, y con sus hijos Don
Sancho y Don Diego llevallos a Carmona, que es una fuerte y rica
villa del Andalucía, y está cerca de Sevilla. Hecho esto, juntó
arrebatadamente su ejército y aprestó su partida para el reino de
Toledo. Llevaba en su campo tres mil hombres de a caballo; pero la
mitad de ellos, ¡mal pecado![383], eran moros, y de quien no se tenía
entera confianza, ni se esperaba que pelearían con aquel brío y
gallardía que fuera necesario. Dícese que al tiempo de su partida
consultó a un moro sabio de Granada, llamado Benagatin, con quien
tenía mucha familiaridad, y que el moro le anunció su muerte por
una profecía de Merlín[384], hombre inglés que vivió antes deste
tiempo, como cuatrocientos años. La profecía contenía estas
palabras: «En las partes de occidente, entre los montes y el mar,
nacerá una ave negra, comedora y robadora, y tal, que todos los
panales del mundo querrá recoger en sí, y todo el oro del mundo
querrá poner en su estómago, y después gormarlo ha[385], y tornará
atrás. Y no perecerá luego por esta dolencia, caérsele han las
péñolas, y sacarle han las plumas al sol, y andará de puerta en
puerta y ninguno la querrá acoger, y encerrarse ha en la selva y allí
morirá dos veces: una al mundo, y otra a Dios, y desta manera
acabará.» Esta fué la profecía, fuese verdadera o ficción, de un
hombre vanísimo que le quisiese burlar; como quiera que fuese, ella
se cumplió dentro de muy pocos días.
El Rey Don Pedro, con la hueste que hemos dicho, bajó del
Andalucía a Montiel, que es una villa en la Mancha y en los Oretanos
antiguos, cercada de muralla, con su pretil, torres y barbacana,
puesta en un sitio fuerte y fortalecida con un buen castillo. Sabida
por Don Enrique la venida de Don Pedro, dejó a Don Gómez
Manrique, Arzobispo de Toledo, para que prosiguiese el cerco de
aquella ciudad, y él, con dos mil y cuatrocientos hombres de a
caballo, por no esperar el paso de la infantería, partió con gran
priesa en busca de Don Pedro. Al pasar por la villa de Orgaz, que
está a cinco leguas de Toledo, se juntó con él Beltrán Claquin[386] con
seiscientos caballos extranjeros que traía de Francia; importantísimo
socorro y a buen tiempo, porque eran soldados viejos y muy
ejercitados y diestros en pelear. Llegaron al tanto[387] allí Don
Gonzalo Mejía, maestre de Santiago, y Don Pedro Muñiz[388], maestre
de Calatrava, y otros señores principales que venían con deseo de
emplear sus personas en la defensa y libertad de su patria. Partió
Don Enrique con esta caballería; caminó toda la noche, y al
amanecer dieron vista a los enemigos, antes que tuviesen nuevas
ciertas que eran partidos de Toledo.
Ellos, cuando vieron que estaba tan cerca Don Enrique, tuvieron
gran miedo, y pensaron no hobiese alguna traición y trato para
dejarlos en sus manos; a esta causa[389] no se fiaban los unos de los
otros. Recelábanse también de los mismos vecinos de la villa. Los
capitanes, con mucha priesa y turbación, hicieron recoger los más
de los soldados que estaban alojados en las aldeas cerca de Montiel;
muchos dellos desampararon las banderas de miedo o por el poco
amor y menos gana con que servían.
Al salir del sol formaron sus escuadrones de ambas partes y
animaron sus soldados a la batalla. Don Enrique habló a los suyos en
esta sustancia[390]: «Este día, valerosos compañeros, nos ha de dar
riquezas, honra y reino, o nos lo ha de quitar. No nos puede suceder
mal, porque de cualquiera manera que nos avenga, seremos bien
librados; con la muerte, saldremos de tan inmensos e intolerables
afanes como padecemos; con la victoria, daremos principio a la
libertad y descanso, que tanto tiempo ha deseamos. No podemos
entretenernos ya más; si no matamos a nuestro enemigo, él nos ha
de hacer perecer de[391] tal género de muerte, que la ternemos[392]
por dichosa y dulce si fuere ordinaria, y no con crueles y bárbaros
tormentos. La naturaleza nos hizo gracia de la vida con un necesario
tributo, que es la muerte; ésta no se puede excusar; empero los
tormentos, las deshonras, afrentas e injurias, evitarálas vuestro
esfuerzo y valor. Hoy alcanceréis una gloriosa victoria, o quedaréis
como honrados y valerosos tendidos en el campo. No vean tal mis
ojos; no permita vuestra bondad, Señor, que perezcan tan virtuosos
y leales caballeros. Mas ¿qué muerte tan desastrada y miserable nos
puede venir que sea peor que la vida acosada que traemos? No
tenemos guerra con enemigo que nos concederá partidos
razonables, ni aun una tolerable servidumbre, cuando queramos
ponernos en sus manos; ya sabéis su increíble crueldad, y tenéis
bien a vuestra costa experimentado cuán poca seguridad hay en su
fe y palabra. No tiene mejor fiesta, ni más alegre[393], que la que
solemniza con sangre y muertes, con ver destrozados los hombres
delante de sus ojos. ¿Por ventura habémoslo[394] con algún malvado
y perverso tirano, y no con una inhumana y feroz bestia, que parece
ha sido agarrochada en la leonera para que de allí con mayor
braveza salga a hacer nuevas muertes y destrozos? Confío en Dios, y
en su apóstol Santiago, que ha caído en la red que nos tenía tendida
y que está encerrado, donde pagará la cruel carnicería que en
nos[395] tiene hecha; mirad, mis soldados, no se os vaya; detenedla,
no la dejéis huir, no quede lanza ni espada que no pruebe en ella sus
aceros. Socorred, por Dios, a nuestra miserable patria, que la tiene
desierta y asolada; vengad la sangre que ha derramado de vuestros
padres, hijos, amigos y parientes. Confiad en nuestro Señor, cuyos
sagrados ministros sacrílegamente ha muerto, que os favorecerá
para que castiguéis tan enormes maldades, y le hagáis un agradable
sacrificio de la cabeza de un tal monstruo horrible y fiero tirano»[396].
Acabada la plática, luego con gran brío y alegría arremetieron a
los enemigos; hirieron en ellos con tan gran denuedo, que sin poder
sufrir este primer ímpetu en un momento fueron desbaratados. Los
primeros huyeron los moros[397], los castellanos resistieron algún
tanto; mas como se viesen perdidos y desamparados, se recogieron,
con el Rey Don Pedro, en el castillo de Montiel. Murieron muchos de
los moros en la batalla, muchos más fueron los que perecieron en el
alcance[398]; de los cristianos no murió sino sólo un caballero[399].
Ganóse esta victoria un miércoles, catorce días de marzo del año de
1369.
Don Enrique, visto cómo Don Pedro se encerró en la villa, a la
hora la hizo cercar de una horma (pared de piedra seca) con gran
vigilancia porque no se les pudiese escapar. Comenzaron los
cercados a padecer falta de agua y de trigo, ca lo poco que tenían
les dañó de industria[400], a lo que parece, algún soldado de los de