Development and application of techniques for quantitative voxel based morphometry [Elektronische Ressource] / vorgelegt von Veronika Ermer

Veronika Ermer

Sujets

• Gesundheit

Informations

• Publié par: rheinisch-westfalischen_technischen_hochschule_-rwth-_aachen
• Publié le: 01 janvier 2008
• Nombre de lectures: 20
• Langue: English
• Poids de l'ouvrage: 11 Mo

Extrait

Development and Application of
Techniques for Quantitative Voxel-Based
Morphometry








Von der Medizinischen Fakultät der Rheinisch-Westfälischen Technischen
Hochschule Aachen zur Erlangung des akademischen Grades einer Doktorin der
Theoretischen Medizin genehmigte Dissertation


vorgelegt von



Dipl.-Ing. (FH) und Master of Science in Biomedical Engineering


Veronika Ermer


aus München




Berichter: Herr Universitätsprofessor
Dr. rer. nat. N. Jon Shah

Herr Universitätsprofessor
Dr. rer. nat. Klaus Willmes-von Hinckeldey



Tag der mündlichen Prüfung: 19. November 2008



Diese Dissertation ist auf den Internetseiten der Hochschulbibliothek
online verfügbar.












The master in the art of living makes little distinction between
her work and her play, her labor and her leisure,
her mind and her body,
her education and her recreation, her love and her religion.
She hardly knows which is which.
She simply pursues her vision of excellence
in whatever she does, leaving others to decide
whether she is working or playing.

To her she is always doing both.




Buddhist Text
ii Abstract

ABSTRACT


The use of standard brain atlases is well established in the MR community, but
none of the commonly utilised standard brains or atlases, such as MNI305 or
ICBM152, provide quantitative information. Within the framework of quantitative
MRI of the brain, which has gained much importance in the last few years, this
work reports on the development of the first quantitative brain atlas for tissue
water content.

A fast and reliable method for quantitatively measuring the absolute water content
of the brain, developed previously in-house, was used and sequence protocols
were optimised to achieve the highest precision and accuracy of the individual
*measurements. Water content, T and T were simultaneously mapped based on 1 2
*an acquired series of spoiled gradient echo images with different T -weighting 2
(QUTE). Based on the quantitative information of these maps, three brain atlases
were developed, representative for a young and healthy population with an
average age of 33.5 + 12.07 years. Data were transformed into a common
stereotactic space and linearly averaged to form the final atlases. Several
approaches for atlas formation were extensively investigated and validated.

Subsequently, the quantitative water content atlas was used as reference for the
comparison of patients with pathological changes in the absolute water content of
the brain to healthy volunteers. A voxel-wise comparison to the atlas was
performed, using quantitative water content maps of patients suffering from
multiple sclerosis (MS) and patients undergoing electro-convulsive therapy (ECT).
Comparing water content maps of MS patients to the reference water content
atlas, using a one-sample t-test, led to clearly defined WM lesions. During the
course of ECT, small changes in the water content were found.

In summary, the first quantitative water content brain atlas in vivo was developed
in the current work and a voxel-wise comparison of diseases related to a change
in the brain water content was performed. With the results presented and further
improvements, quantitative MRI in combination with a quantitative water content
brain atlas allows for a careful and quantitative interpretation of disease
monitoring.

iii Acknowledgments

ACKNOWLEDGMENTS


Scientifically, but also personally this project has been a real challenge and would
not have been realised with the help of others. It is a pleasure for me to thank the
many people who made this thesis possible.

First and foremost I would like to thank Prof. Dr. N. Jon Shah without whom this
thesis would not exist. He gave me the opportunity to join the MR-group in the
Institute of Neuroscience and Biophysics 3 - Medicine (Research Centre of Jülich)
and to work in this agreeable and inspiring atmosphere. I also want to thank him
for being such a patient supervisor and for supporting this work with ideas,
criticism and a lot of helpful discussions and suggestions.

I would like to especially thank Prof. Dr. Heiko Neeb for his very helpful and patient
support, discussions at any time and finally for his highly accurate and precise
corrections and comments.

Furthermore, I would like to gratefully acknowledge the help of Dr. Tony Stöcker.
Especially, his very helpful discussions on the topic of VBM and all aspects and
functions of SPM supported me through the three years.

Importantly, I want to acknowledge the MTAs of the MR group: Petra Engels,
Barbara Elghahwagi, Dorothe Krug and Cordula Kemper. Their support and help
in volunteer and patient recruitment and the performance of experiments
throughout the whole time was invaluable.

Another thanks to the whole MR group: it was a great experience to be a member
of this group from the first day on.

Last but not least, I want to thank my parents and my brother for all their support
during the last 30 years and particularly for giving me the opportunity to carry on
my studies here in Jülich.
iv Table of Contents


TABLE OF CONTENTS

A) LIST OF FIGURES vii
B) LIST OF TABLES xiii
C) LIST OF VARIABLES xv
D) LIST OF ABBREVIATIONS xvii

1. INTRODUCTION 1
2. THEORY OF NMR 5
2.1. Nuclear Spins and Magnetic Moment 5
2.2. The Quantum Mechanical description of NMR 6
2.3. Boltzmann Statistics 9
2.4. Relaxation Processes 10
2.4.1. Bloch equations 10
2.4.2. Radio Frequency (RF) excitation 11
2.4.3. Relaxation times 12
2.4.3.1. Spin-Lattice (Longitudinal) Relaxation 12
2.4.3.2. Spin-Spin (Transverse) Relaxation 13
2.4.4. Ernst Formula 15
2.4.5. Relaxation and molecular motion 16
2.4.5.1. Microscopic Mechanism 16
2.4.5.2. Influences on the relaxation times 19
3. THEORY OF MRI 21
3.1. K-Space Formalism 22
3.2. Standard sequences 24
3.2.1. Echo-Planar Imaging 24
* 3.2.2. Quantitative T Image (QUTE) 25 2
3.3. Water Content Mapping 27
4. VOXEL-BASED MORPHOMETRY 31
4.1. Approaches to VBM 32
4.2. Segmentation 34
4.3. Normalisation 37
4.4. Smoothing 40
4.5. Statistical Analysis 4
v Table of Contents
5. ATLAS DEVELOPMENT 43
Part I
5.1. Methods & Subjects I 45
5.1.1. Simulations 46
5.1.2. Phantom Measurements 50
5.2. Results I 53
5.2.1. Simulations 53
5.2.2. Phantom Experiments 58
5.3. Discussion & Prospects I 60

Part II
5.4. Methods & Subjects 63
5.4.1. In vivo Experiments 63
5.4.2. Approaches For The Construction of The Quantitative Water
Content Brain Atlas 65
5.5. Results II 69
5.5.1. In vivo Experiments 69
5.5.2. Approaches For The Construction of The Quantitative Water
Content Brain Atlas 72
5.6. Discussion & Prospects II 81
6. APPLICATION TO MS 85
6.1. Subjects & Methods 87
6.2. Results 89
6.3. Discussion & Prospects 93
7. APPLICATION TO ECT 97
7.1. Methods & Subjects 98
7.2. Results 100
7.3. Discussion & Prospects 103
8. FUTURE PROSPECTS 105

REFERENCES 107

ERKLÄRUNG ZUR DATENAUFEBWAHRUNG 113

LEBENSLAUF 115




vi List of Figures



A) LIST OF FIGURES

Figure 1: Zeeman energy levels for a spin 3/2 nucleus in a static magnetic
field, B 7 0

Figure 2: The net magnetisation is moved from the equilibrium position along the
z-axis in accordance with the applied RF pulse. The flip angle, !, depends on
the pulse strength and its duration. 11

Figure 3: Longitudinal relaxation of magnetisation. After excitation using a 90°
pulse, the longitudinal magnetisation relaxes to the equilibrium value M =M . Z 0
The temporal development is characterised by the relaxation time T . 13 1

Figure 4: Relaxation of the transverse magnetisation. After excitation by a 90°
pulse, the transverse magnetisation relaxes to the equilibrium value M =0. XY
The temporal development is characterised by the relaxation time T . 14 2

Figure 5: Schematic dependence of the relaxation times T and T on molecular 1 2
correlation times, ! . A) T >> T : rigid structures; B - C) T > T : non-rigid C 1 2 1 2
solids (e.g. tissue water); D) T = T : non-viscous liquids (free water) (Farrar 1 2
1971). 18

Figure 6: MR pulse sequence diagram of the EPI sequence. A slice-selective RF
pulse is followed by a series of readout gradients in x-direction which are
accompanied by phase-encoding gradients in y-direction to spatially encode
the acquired signal in two dimensions. 25

Figure 7: M