Ebook Microscopic magnetic resonance imaging: Part 1

(BQ) Part 1 book Microscopic magnetic resonance imaging has contents: About this book, hardware, image formation, acquisition strategies, image artifacts, sample preparation,... and other contents.

Microscopic Magnetic Resonance Imaging Pan Stanford Series on Renewable Energy — Volume Microscopic Magnetic Resonance Imaging A Practical Perspective editors Preben Maegaard Anna Krenz Wolfgang Palz Luisa Ciobanu The Rise of Modern Wind Energy Wind Power for the World Published by Pan Stanford Publishing Pte Ltd Penthouse Level, Suntec Tower Temasek Boulevard Singapore 038988 Email: editorial@panstanford.com Web: www.panstanford.com British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Microscopic Magnetic Resonance Imaging: A Practical Perspective Copyright c 2017 Pan Stanford Publishing Pte Ltd All rights reserved This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the publisher Cover: Two neurons of Aplysia californica imaged with MRM For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA In this case permission to photocopy is not required from the publisher ISBN 978-981-4774-71-0 (Paperback) ISBN 978-981-4774-42-0 (Hardback) ISBN 978-1-315-10732-5 (eBook) Printed in the USA June 7, 2017 12:43 PSP Book - 9in x 6in To my parents: Nature or nurture, I owe it all to you 00-MRI-Prelims Contents xi Preface SECTION I INTRODUCTION About This Book SECTION II BASICS Hardware 2.1 The Main Magnet 2.2 Radiofrequency Coils 2.2.1 Basic Coil Designs 2.2.1.1 Solenoidal coils 2.2.1.2 Surface coils 2.2.2 RF Circuit Design 2.2.3 Coil Performance 2.2.4 Other Types of Coils 2.2.4.1 Inductively coupled coils 2.2.4.2 Cryogenically cooled coils 2.3 Gradient Coils 2.4 The Elusive “Key” Component 7 10 12 13 15 17 17 18 19 20 Image Formation 3.1 The Bloch Equation 3.2 The k-space 3.3 Encoding Schemes 3.4 Image Resolution 23 23 24 25 27 viii Contents 3.5 Image Signal-to-Noise-Ratio 3.6 Choice of Imaging Parameters 28 30 Acquisition Strategies 4.1 Pulse Sequences 4.1.1 Spin Echo (SE) 4.1.1.1 Fast spin echo acquisitions 4.1.2 Gradient Echo (GE) 4.1.2.1 Fast gradient echo acquisitions 4.1.3 Hybrid Pulse Sequences 4.1.4 Accelerated Acquisitions 4.2 Contrast Mechanisms 4.2.1 Basic Contrasts 4.2.1.1 MR contrast agents 4.2.2 Diffusion-Weighted Imaging 4.2.2.1 Diffusion acquisitions for MR microscopy 33 33 33 35 36 37 38 39 41 41 43 44 Image Artifacts 5.1 Magnetic Susceptibility Artifacts 5.2 Chemical Shift Artifacts 5.3 Motion Artifacts 5.4 Aliasing Artifacts 5.5 RF Coil Calibration Artifacts 5.6 Clipping Artifacts 5.7 Gibbs Ringing Artifacts 5.8 Zipper Artifacts 5.9 Spurious Echoes Artifacts 51 51 54 55 56 57 58 58 60 60 Sample Preparation 6.1 Fixed Tissues 6.2 Live Tissues 63 63 65 49 SECTION III APPLICATIONS A Bit of History 7.1 Biological Detour: The Aplysia 69 70 Contents 7.1.1 The Buccal Ganglia 7.1.2 The Abdominal Ganglion 7.2 Advances in Spatial Resolution 72 73 73 Diffusion Weighted Magnetic Resonance Microscopy 8.1 Diffusion and Tissue Microstructure 8.2 Diffusion and Neuronal Function 81 81 85 Manganese Enhanced Magnetic Resonance Microscopy 9.1 In vivo Manganese Administration 9.2 Ex vivo Manganese Administration 9.3 Manganese Toxicity 89 90 94 95 SECTION IV CONCLUSION 10 On the Horizon 101 Appendix References Index 103 107 117 ix 48 Acquisition Strategies (a) 0.001 (b) 0 Figure 4.12 (a) FA and (b) MD maps of a chemically fixed mouse brain slice Spatial resolution: 50 μm × 50 μm × 130 μm Operating frequency 730 MHz The acquisition parameters are listed in the Appendix or hypointense depending on the direction of the diffusion encoding gradients It is clear that in the presence of anisotropy, diffusion cannot be characterized by a single scalar coefficient; instead it requires a tensor and therefore measurements have to be performed along several gradient directions (minimum six) From the diffusion tensor components, one can calculate different metrics providing information about tissue microarchitecture, the most common ones being mean diffusivity (MD) and fractional anisotropy (FA) (Le Bihan, 2001) Figure 4.12 shows FA and MD maps of a fixed mouse brain The white matter structures can be clearly visualized on the FA map For high b-values, the decay of the diffusion-weighted signal deviates from the mono-exponential model described by Eq 4.17 Empirically this decay was shown to be better fitted by a weighted sum of two exponential functions: Sbiexp = S0 (w1 e−bD1 + w2 e−bD2 ) (4.22) Equation 4.22 suggests the presence of two compartments, characterized by two different diffusion coefficients, D1 and D2 , and for which the weights, w1 and w2 , represent the relative volume fractions The initial interpretation of this model was that the two compartments correspond to the intra and extracellular spaces This hypothesis was, however, not confirmed experimentally and there Contrast Mechanisms is still a debate regarding the physical origins of the two pools Despite this, the biexponential description of tissue water diffusion continues to be used due to its robustness and simplicity Other models have nonetheless been proposed including the kurtosis model (Jensen, 2005), the statistical model (Yablonskiy, 2003) and the stretched exponential model (Bennett, 2003) Among these, the most popular one is the kurtosis model in which the decay of the signal is fitted with a Taylor expansion truncated at the second term: Sbiexp = S0 e−bD+ (bD) K (4.23) In Eq 4.23, K , the kurtosis, quantifies the deviation from the Gaussian behavior (for Gaussian diffusion K = 0) 4.2.2.1 Diffusion acquisitions for MR microscopy Diffusion measurements are time consuming as they typically require multiple b-values and multiple diffusion encoding directions For this reason, the acquisition strategy most commonly used in clinical and preclinical imaging is the diffusion-weighted EPI However, as discussed earlier, EPI acquisitions are not favorable to magnetic resonance microscopy studies due low SNR and their inherent sensitivity to B0 inhomogeneities Incorporating diffusionweighted modules in fast GE acquisitions such as FISP (fast imaging with steady-state free precession) are good alternatives to EPI Such approaches are not suitable for diffusion quantification because the resulting signal is a complicated function of diffusion, T1 and T2 weightings (Deoni, 2004; Le Bihan, 1988) For straightforward diffusion quantification one can use diffusion preparation modules followed by fast acquisitions In the schematic presented in Fig 4.13, the diffusion weighting is imparted to the [ DP 90° 180° 90° δ G diff G diff Δ Figure 4.13 G spoil ][ IMAGING SEQUENCE {FLASH, FISP] ] Schematic of a diffusion preparation module 49 50 Acquisition Strategies Figure 4.14 (a) Axial DP-FISP image of an Aplysia neuron for b = s/mm2 and (b) the corresponding ADC parametric map Spatial resolution 25 μm isotropic Operating frequency 730 MHz The cytoplasm and the nucleus are distinguishable in both images The acquisition parameters are listed in the Appendix Images courtesy of Dr Ileana Jelescu longitudinal magnetization in the preparation module (DP) and the image encoding is performed using fast gradient echo-based acquisitions The main drawback of this preparation scheme is the longitudinal magnetization recovery between the end of the preparation module and the beginning of the acquisition which leads to T1 contamination of the diffusion-weighted signal In order to minimize this contamination, one has to encode the central k-space points immediately after the diffusion preparation by performing centrically ordered encoding Jelescu et al (2014) report the implementation of a 3D diffusion prepared (DP)-FISP pulse sequence with suitable timings (9 per b-value) and resolutions (25 μm isotropic) for MR microscopy of unfixed biological tissue Using this sequence the authors report ADC maps of single isolated neurons in which the nucleus and the cytoplasm can be clearly differentiated (see Fig 4.14) It is worth noting that in order to avoid T1 contamination and to ensure a correct assessment of the effective b-value, this sequence was used for low b-values (