Diffusion Tensor Imaging

Transcription

1

2 Stieltjes Brunner Fritzsche Laun Diffusion Tensor Imaging Introduction and Atlas 3-D dataset on CD

3 Diffusion Tensor Imaging

4 Bram Stieltjes Romuald M. Brunner Klaus H. Fritzsche Frederik B. Laun Diffusion Tensor Imaging Introduction and Atlas With 340 figures and CD-ROM 123

5 Dr. Bram Stieltjes Deutsches Krebsforschungszentrum (DKFZ) Section Quantitative imaging-based disease characterization, department of radiology Im Neuenheimer Feld Heidelberg Germany Prof. Dr. Romuald M. Brunner Heidelberg University Hospital, Department of psychiartry, Division child- and adolescent psychiatry Blumenstr Heidelberg Germany Dr. Klaus H. Fritzsche Deutsches Krebsforschungszentrum (DKFZ) Department of medical and biological informatics Im Neuenheimer Feld Heidelberg Germany Dr. Frederik B. Laun Deutsches Krebsforschungszentrum (DKFZ) Department of medical physics in radiology Im Neuenheimer Feld Heidelberg Germany Additional material to this book can be downloaded from ISBN DOI / ISBN (ebook) Springer Medizin Springer-Verlag Berlin Heidelberg 2013 Library of Congress Control Number: This work is subject to copyright. All rights are reserved 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. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. 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. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Editor: Dr. Christine Lerche Project Management: Claudia Bauer Copyediting: Isabella Athanassiou, Heidelberg Project Coordination: Barbara Karg Cover Illustration: Dr. Bram Stieltjes, Deutsches Krebsforschungszentrum, Heidelberg Cover Design: deblik Berlin Typesetting and Reproduction of the figures: Fotosatz-Service Köhler GmbH Reinhold Schöberl, Würzburg The German Cancer Research Center (Deutsches Krebsforschungszentrum) holds the rights to the software on the CD. Printed on acid-free paper Springer Medizin is brand of Springer Springer is part of Springer Science+Business Media (

6 V Preface Since the advent of diffusion tensor imaging in the mid-1990s, the field of in vivo representation of human neuronal connectivity has experienced a dramatic development in terms of technological advances as well as of applications in both neuroscience and clinical research. Although the main focus of this book is on the depiction of neuroanatomy derived from diffusion imaging, we hope that with the»introduction to Diffusion Imaging,«the book will also be a good first point of reference for navigation of the currently overwhelmingly extensive literature on diffusion imaging. The making of this atlas was a far more interactive process than initially anticipated. For instance, to be able to create reproducible image views and constant size representations, we had to redesign our fiber tracking software and in many ways, over time, the final result evolved far from the original concept. We hope that the resulting combination of two-dimensional T 1 -weighted and color map-based anatomy enriched with the part on three-dimensional white matter tract representation will help readers to quickly find their way in the most fascinating of all mazes, the human brain. Finally, we want to thank T. Kuder for corrections and graph preparation and Springer Heidelberg for embarking with us on this adventurous journey; Mrs. R. Scheddin for the enthusiasm when we first presented the idea of the atlas and Dr. C. Lerche and Mrs. C. Bauer for unrelenting support during the numerous iterations and corrections. Bram Stieltjes Romuald M. Brunner Klaus H. Fritzsche Frederik B. Laun Spring 2012

7 VII Contents I Introduction How to Use this Atlas Introduction to Diffusion Imaging Diffusion: A Primer Timeline Theoretical Aspects Advanced Techniques: Fiber Tracking and Deviations from Mono-exponential Signal Decay Practical Aspects Selected Applications in Neuroscience References II Atlas 2 Two-dimensional Brain Slices Coronal View Sagittal View Transversal View Three-dimensional Fiber Tracking Fiber Tracking of the Cerebral Hemispheres Fiber Tracking of the Brain Stem III Appendix Index Introduction Index Atlas

8 1 I Section I Introduction How to Use this Atlas 3 Chapter 1 Introduction to Diffusion Imaging 5

9 How to Use this Atlas 3

10 4 How to Use this Atlas jthe Timeline and Introduction The timeline and introduction sections can be seen as a primer in diffusion imaging. Although the atlas focuses on directional diffusion, we chose to give an outline of the complete diffusion literature to come to a comprehensive overview. To make the introduction accessible to readers with a different academic background, we designed it so that the main body of the text is understandable without specialist knowledge in physics or mathematics. More detailed information on the diffusion process and calculations can be found in separate textboxes for further reading. Thus, as your knowledge of diffusion increases, so may your interest in the content of the textboxes. jthe Atlas The atlas consists of two major parts: a two- and a three-dimensional representation of fiber tracts. In the two-dimensional part, we compare the conventional T 1 -weighted anatomy with a diffusion tensor imaging-derived color map. In these maps, the directional orientation of fiber tracts is color coded in the following fashion: tracts moving leftright are coded red (e.g., the corpus callosum), anterior-posterior tracts are coded green (e.g., the cingulum), and craniocaudal tracts are coded blue (e.g., the corticospinal tract). The intensity or hue indicates the fractional anisotropy, a measure of fiber density. The whole brain is covered in the three main radiological planes: axial, coronal, and sagittal. The three-dimensional part covers the most prominent white matter connections in the human brain. The complete reconstruction process is presented in a consistent, step-by-step fashion. First, the relevance and anatomy of the tract are discussed and an initial region of interest (ROI) is shown. This ROI is chosen to yield an optimal final result and is shown in white. By using inclusion (green) and exclusion (red) ROIs, the result is further refined. The final result is represented without the ROIs in three different planes as well as in an oblique view for optimal appreciation of the anatomical location. The color coding of these tracts is identical to the two-dimensional color maps. The intricate anatomy of important adjacent tracts is further illustrated in combined overviews. Here, each individual tract is represented in monochrome to enhance the visualization of the complex interwoven anatomy. Again, these results are represented in three standard planes and an oblique view. All tracts in the three-dimensional part are also indicated in the two-dimensional part of the atlas, and flipping between these parts will help readers to understand the projection of the threedimensional tract on the two-dimensional slices. Using the T 1 -weighted images, adjacency to important gray matter structures is captured. Thus, by interactively using the atlas, your understanding of the complex neuroanatomical correlations should grow continuously. To aid this process, the accompanying CD contains all the tracts as illustrated in the atlas and allows you to scroll through the data in an interactive fashion.

11 5 1 Introduction to Diffusion Imaging 1.1 Diffusion: A Primer Timeline Theoretical Aspects Advanced Techniques: Fiber Tracking and Deviations from Mono-exponential Signal Decay Practical Aspects Selected Applications in Neuroscience References 37 B. Stieltjes et al., Diffusion Tensor Imaging, DOI / _1, Springer-Verlag Berlin Heidelberg 2013

12 6 Chapter 1 Introduction to Diffusion Imaging Diffusion: A Primer Diffusion is for the purpose of this book the random motion of water molecules in fluid water. When water is frozen, the particles stand still, but when water is liquid, the water molecules move around owing to their thermal energy. They do so rather quickly, at about 1,000 m/s. This above-racing-car speed, however, does not convert into a high-speed, long-distance distribution of the molecules, because they bounce at each other very frequently: The time between bounces is about 1/1,000,000,000,000 s only. Despite this high-speed crashing which, of course, prevents the particle from effectively translating like a racing car the particles do move away from their original position. Since this process depends so crucially on the exact initial positions and velocities of all the 10,000,000,000,000,000,000,000 water molecules present in each drop of water (which is information that is totally elusive to us), it appears to us that the water molecules move in a random fashion making it impossible to predict where they end up. Thus, because the system is so incredibly complex, the human investigator does not care about the individual water molecule, but only about Joe Water. How far does Joe Water translate? Of course, that depends on the time we let him bounce around with his fellows, but let us say: How far does he translate in the time (approximately 0.1 s) we need to make a magnetic resonance (MR) image of the head where Joe is supposed to stay momentarily. The answer, which was verified in a variety of experiments, is: he moves about 10 μm. Great! The average cell has approximately the same size. Joe will most likely bounce against the cell membrane. Thus, already slowed down by his fellows, Joe has found another obstacle curbing his freedom. And this obstacle, although annoying for the pert and nimble spirit of Joe, means a lot to us. For it is thought (by many and with some entitlement) that if we have information about the diffusion-restricting structure of the cell membrane, then we will know more about the tissue, about its diseases, and about its course, if it is an axonal tract, thereby enlightening us about its orientation, be it left right, up down, or top bottom. Thus, if we could measure diffusion, it would open up a new world for us, providing a glance at microcellular structures, allowing apparently straightforward interpretations to be made. Take a drug to fight the tumor. Diffusion goes up? Good, the cell membrane gets broken! Or: Diffusion is largest in left right direction? There must be a nerve bundle! If only we were able to measure diffusion in a pain-free, convenient fashion... And, indeed, we can! This might come as no surprise, for otherwise you would probably not be reading this primer of an atlas on MR diffusion tensor imaging. This might also come as no surprise since one knows that MR imaging yields images of the spatial distribution of water molecules; so the step of measuring displacements does not seem to be too far off. In fact, diffusion is in many regards measured like an MR image. Apart from acquiring one image, one designs an experiment that intrinsically acquires two images one at the beginning of the experiment, and one at the end that are subtracted from each other. If Joe was lazy and did not move, the two images would cancel, but if he moved from A to B, then the images would not cancel, and from the difference image, we could learn about the separation of A and B. In practice, of course, things are slightly more complicated: it is not really an image that is acquired, but a phase image. Moreover, it is not possible to measure the difference image of one particle, but only that of the whole ensemble, which renders the process more complex. This complicated topic is explained in the first part of this atlas. After a short historical overview in 7 Sect. 1.2, the required sophisticated techniques are explained in 7 Sect. 1.3 to 7 Sect We put much effort into rendering the manuscript accessible to a wide readership. Since most readers might be happy to get the picture without being bothered by forests of formulas, we designed the main text such that it should be understandable without a PhD in physics. Nonetheless, since diffusion is not an easy topic, it might require duty and desire to get through it. The text is written assuming that the reader is familiar with basic mathematical tools (addition, subtraction, multiplication, division, exponentials, and logarithms) and basic MR imaging concepts (T 1, T 2, Larmor frequency, rotating reference frame, spin echo, k-space), although much of the text should be comprehensible without these prerequisites. Current or future PhD physicists should also enjoy 7 Sect. 1.3 to 7 Sect. 1.6, since there are special textboxes, full of formulas, presenting an in-depth treatment of the topic.

13 1.2 Timeline Timeline (. Fig. 1.1,. Fig. 1.2,. Fig. 1.3). Fig. 1.1 Main developments in diffusion NMR imaging:

14 8 Chapter 1 Introduction to Diffusion Imaging 1. Fig. 1.2 Main developments in diffusion NMR imaging:

15 1.2 Timeline 9 1. Fig. 1.3 Main developments in diffusion NMR imaging: