L’imagerie en tenseur de diffusion, ou DTI, est une application de l’imagerie de diffusion qui permet de quantifier en chaque direction de l’espace la diffusion des molécules d’eau. Cette technique permet d’obtenir la direction de fibres cérébrales en chaque voxel, et de reconstruire indirectement les faisceaux de substance blanche du cerveau en 3D par tractographie. Les paramètres scalaires du tenseur, la FA ou fraction d’anisotropie, et l’ADC ou coefficient apparent de diffusion, permettent d’analyser la microstructure cérébrale de manière quantifiée. Les applications du DTI sont nombreuses, comme l’étude du développement cérébral normal et des pathologies de la substance blanche.Nous avons tout d’abord étudié le DTI chez le fœtus. Pour ce faire, une chaîne de traitement d’images DTI fœtales, compilée dans un logiciel, Baby Brain Toolkit (BTK) (https://github.com/rousseau/fbrain), a été implémentée. Ce logiciel permet notamment de corriger les artéfacts de mouvements qui dégradent la qualité du DTI fœtal. BTK a été validé sur des cas normaux, puis a été appliqué à un modèle de malformation cérébrale. Nous avons aussi étudié un cas d’infection à cytomégalovirus en DTI.Nous avons ensuite analysé l’intérêt des paramètres scalaires DTI dans l’étude d’une leucodystrophie rare, le syndrome de Cockayne. Le DTI permet de diagnostiquer le syndrome de Cockayne, de distinguer ses sous-types cliniques, et d’approcher sa physiopathologie. Nous avons ainsi montré qu’il s’agit d’une pathologie hypomyélinisante primitive, suivie d’une démyélinisation secondaire de bas grade. / Diffusion tensor imaging (DTI) is a diffusion-weighted imaging application that allows water motion quantification in any direction. This technique determines brain fiber direction in each voxel, and reconstructs indirectly white matter fibers tracts in 3D with tractography. Scalar DTI parameters, such as fractional anisotropy (FA) and apparent diffusion coefficient (ADC), provide a quantitative analysis of brain microstructure. DTI applications are numerous, especially in the study of brain development and white matter pathologies.First, we studied DTI in the fetus. For this, we implemented a processing method for fetal DTI images, and compiled it in a software, Baby brain Toolkit (BTK) (https://github.com/rousseau/fbrain). BTK was validated on normal cases, and then applied to a brain malformation model. We also studied a case of cytomegalovirus infection with DTI.We then investigated the utility of scalar DTI parameters in a rare leukodystrophy, Cockayne syndrome. DTI allows to diagnose Cockayne syndrome, to distinguish between clinical subtypes, and to understand its pathophysiology. We showed that Cockayne syndrome was a primitive hypomyelinating disorder, followed by a low grade secondary demyelination.

Les outils méthodologiques en indexation et classification des images par le contenu sont déjà assez matures et ce domaine s’ouvre vers les applications médicales. Dans cette thèse,nous nous intéressons à l'indexation visuelle, à la recherche et à la classification des images cérébrales IRM par le contenu pour l'aide au diagnostic de la maladie d'Alzheimer (MA). L'idée principale est de donner au clinicien des informations sur les images ayant des caractéristiques visuelles similaires. Trois catégories de sujets sont à distinguer: sujets sains (NC), sujets à troubles cognitifs légers (MCI) et sujets atteints par la maladie d'Alzheimer(AD). Nous représentons l’atrophie cérébrale comme une variation de signal dans des images IRM (IRM structurelle et IRM de Tenseur de Diffusion). Cette tâche n'est pas triviale,alors nous nous sommes concentrés uniquement sur l’extraction des caractéristiques à partir des régions impliquées dans la maladie d'Alzheimer et qui causent des changements particuliers dans la structure de cerveau : l'hippocampe le Cortex Cingulaire Postérieur. Les primitifs extrais sont quantifiés en utilisant l'approche sac de mots visuels. Cela permet de représenter l’atrophie cérébrale sous forme d’une signature visuelle spécifique à la MA.Plusieurs stratégies de fusion d’information sont appliquées pour renforcer les performances de système d’aide au diagnostic. La méthode proposée est automatique (sans l’intervention de clinicien), ne nécessite pas une étape de segmentation grâce à l'utilisation d'un Atlas normalisé. Les résultats obtenus apportent une amélioration par rapport aux méthodes de l’état de l’art en termes de précision de classification et de temps de traitement. / Content-Based Visual Information Retrieval and Classification on Magnetic Resonance Imaging (MRI) is penetrating the universe of IT tools supporting clinical decision making. A clinician can take profit from retrieving subject’s scans with similar patterns. In this thesis, we use the visual indexing framework and pattern recognition analysis based on structural MRIand Tensor Diffusion Imaging (DTI) data to discriminate three categories of subjects: Normal Controls (NC), Mild Cognitive Impairment (MCI) and Alzheimer's Disease (AD). The approach extracts visual features from the most involved areas in the disease: Hippocampusand Posterior Cingulate Cortex. Hence, we represent signal variations (atrophy) inside the Region of Interest anatomy by a set of local features and we build a disease-related signature using an atlas based parcellation of the brain scan. The extracted features are quantized using the Bag-of-Visual-Words approach to build one signature by brain/ROI(subject). This yields a transformation of a full MRI brain into a compact disease-related signature. Several schemes of information fusion are applied to enhance the diagnosis performance. The proposed approach is less time-consuming compared to the state of thearts methods, computer-based and does not require the intervention of an expert during the classification/retrieval phase.

Cette thèse propose des techniques pour le traitement d'images IRM de diffusion. Les méthodes proposées concernent l'estimation et la régularisation, le groupement et la segmentation ainsi que le recalage. Le cadre variationnel proposé dans cette thèse pour l'estimation d'un champ de tenseurs de diffusion à partir d'observations bruitées exploite le fait que les données de diffusion représentent des populations de fibres et que chaque tenseur peut être reconstruit à partir d'une combinaison pondérée de tenseurs dans son voisinage. La méthode de segmentation traite aussi bien les voxels que les fibres. Elle est basée sur l'utilisation de noyaux défini-positifs sur des probabilités gaussiennes de diffusion afin de modéliser la similarité entre tenseurs et les interactions spatiales. Ceci permet de définir des métriques entre fibres qui combinent les informations de localisation spatiale et de tenseurs de diffusion. Plusieurs approches de groupement peuvent être appliquées par la suite pour segmenter des champs de tenseurs et des trajectoires de fibres. Un cadre de groupement supervisé est proposé pour étendre cette technique. L'algorithme de recalage utilise les noyaux sur probabilités pour recaler une image source et une image cible. La régularité de la déformation est évaluée en utilisant la distortion induite sur les distances entre probabilités spatialement voisines. La minimisation de la fonctionnelle de recalage est faite dans un cadre discret. La validation expérimentale est faite sur des images du muscle du mollet pour des sujets sains et pour des patients atteints de myopathies. Les résultats des techniques développées dans cette thèse sont encourageants. / In this thesis, we present several techniques for the processing of diffusion tensor images. They span a wide range of tasks such as estimation and regularization, clustering and segmentation, as well as registration. The variational framework proposed for recovering a tensor field from noisy diffusion weighted images exploits the fact that diffusion data represent populations of fibers and therefore each tensor can be reconstructed using a weighted combination of tensors lying in its neighborhood. The segmentation approach operates both at the voxel and the fiber tract levels. It is based on the use of Mercer kernels over Gaussian diffusion probabilities to model tensor similarity and spatial interactions, allowing the definition of fiber metrics that combine information from spatial localization and diffusion tensors. Several clustering techniques can be subsequently used to segment tensor fields and fiber tractographies. Moreover, we show how to develop supervised extensions of these algorithms. The registration algorithm uses probability kernels in order to match moving and target images. The deformation consistency is assessed using the distortion induced in the distances between neighboring probabilities. Discrete optimization is used to seek an optimum of the defined objective function. The experimental validation is done over a dataset of manually segmented diffusion images of the lower leg muscle for healthy and diseased subjects. The results of the techniques developed throughout this thesis are promising.

Frontotemporal lobar degeneration (FTLD) is a heterogeneous group of illnesses which can be difficult to diagnose. Modern diagnostic criteria require the presence of imaging abnormalities, but these are not always seen in the early stages of the illness. Hence there is a need to consider the use of more advanced MR techniques. This thesis reports the results of a multimodal MRI study of patients with FTLD, and considers two things: how well data from the different modalities can classify patients, and how well the different modalities can identify affected tissue. FTLD is thought to involve alterations in cerebral blood flow, but it is possible that microvascular changes will alter additional perfusion parameters, such as the time taken for blood to reach the tissue (the arrival time). Multi-time point arterial spin labelling (ASL) measurements have the ability to extract the relevant parameters. I consider the parameters involved in modelling these data, and report the accuracy of cerebral blood flow (CBF) measurement achievable in a clinically acceptable time. FTLD patients have atrophy in the frontal and temporal lobes, regions problematic for MRI because of susceptibility artefacts caused by adjacent air spaces. I consider two ASL MR read-out sequences (gradient-echo and spin-echo)and show that spin-echo images give higher signal in frontal and temporal regions than gradient-echo. ASL, T1-weighted and diffusion-weighted images were collected for a group of 17 FTLD patients and 18 controls. I found decreased CBF in highly atrophied regions of cortical grey matter in patients, but this deficit was not seen when corrected for atrophy. An increased arrival time was seen in regions adjacent to the atrophied regions, but a decreased arrival time was seen in the atrophied regions; this is a novel finding. The diffusion metrics of fractional anisotropy (FA) and particularly mean diffusivity (MD) are found to be highly sensitive to differences in FTLD patients. I speculate that this is an increased sensitivity to atrophy because of the increased signal from cerebrospinal fluid. I combine the regional values of all the modalities in a classification method to distinguish patients from controls, and establish a combination of region and modality that classified 21/22 subjects correctly. This exploratory study is the first time all three modalities have been combined in a study of FTLD patients; it shows that combining MR modalities may lead to improved classification of FTLD patients and better identification of affected tissue.

Glioblastoma multiform (GBM) is an aggressive primary brain tumour. Despite treatment advances in recent years, outcomes remain poor. Disease progression tends to occur adjacent to the original tumour or surgical resection bed, usually within the radiotherapy planning field. This local recurrence and progression is believed to be the result of invasive disease in the surrounding tissue at the time of diagnosis and treatment, which is not currently detectable by conventional non-invasive methods. A number of novel therapies are currently under development which target specific aspects of the tumour behaviour, to try and improve outcomes from this devastating disease. Imaging biomarkers are under development, therefore, in order to provide a non-invasive assessment of tumour extent and behaviour, to provide bespoke image-guided therapies, and detect recurrence or treatment failure at the earliest opportunity. These are also of value in the context of novel therapeutics, which may have a very specific affect on an aspect of tumour behaviour that is not readily apparent on standard clinical imaging. Key to the progression of GBM is the invasion into surrounding white matter. This is followed by a period of tumour growth and subsequent angiogenesis in which microvasculature is produce that is distinct from the highly regulated blood-brain barrier. This thesis covers the development of suite of advanced magnetic resonance imaging (MRI) techniques aimed at characterising those very traits of GBM responsible for the aggressiveness and treatment resistance. Repeatability studies are undertaken to determine the performance of the biomarkers in healthy tissues, and also in a range of gliomas. Dynamic Contrast Enhanced (DCE-) and dynamic susceptibility-enhanced (DSC-)MRI are used to provide estimates of perfusion and permeability in the tumour. In order to address the reasons behind preferential invasion of the corpus callosum, they are used in conjunction with ASL to non-invasively map perfusion territories and watershed regions in the brain through perfusion timing parameters. Diffusion Tensor Imaging (DTI) and quantitative magnetisation transfer (qMT) are used to provide complementary information about white matter integrity, in order to identify changes occurring with glioma invasion as early as possible and direct image-guided treatments at presentation. Their complementary nature is assessed by comparing the two parameters simultaneously in white matter. Additionally, one of the qMT parameters which may be related to tissue pH is shown to be sensitive and specific for the detection of high-grade tumour tissue. Finally, a novel multiparametric imaging biomarker is introduced. Tumour surface mapping assesses the boundary between the solid tumour and surrounding tissue in order to identify areas of potential aggressiveness and invasion. Multiple imaging parameters can be combined to improve specificity and sensitivity. Using the diffusion-weighted imaging parameter, mean diffusivity (MD - also referred to as the apparent diffusion coefficient (ADC)), it is shown to be predictive of clinical outcome in a retrospective and prospective study, while a multiparametric example is given indicating the utility as a predicative biomarker for regions of progression and recurrence, and as potential spatial discriminator for image-guided therapies.

L'imagerie par tenseur de diffusion (DTI) est une technique qui permet d'étudier l'organisation et l'état structurel des faisceaux de substance blanche. L'étude de l'intégrité des faisceaux peut aider à comprendre et à déterminer la sévérité du pronostic des patients après un accident ischémique cérébral (AIC). Dans cette thèse, je présente deux études ; dans la première nous avons comparé les résultats obtenus à partir d'un template du faisceau Corticospinal (FCS) à ceux obtenus par la tractographie chez des sujets sains et des patients post-AIC. Dans les deux groupes, les valeurs de la fraction d'anisotropie (FA) obtenues avec la tractographie étaient plus élevées que celles du template. Cependant, chez les patients, les deux méthodes ont détecté une diminution des valeurs de FA du FCS ipsilésionnel, qui était corrélée aux scores moteurs, mais les valeurs de FA obtenues avec le template étaient mieux corrélées avec le signal BOLD généré par la main parétique. La deuxième étude cherchait à savoir si la sévérité de l'aphasie post-AIC pouvait être corrélée à l'atteinte de certaines régions de la substance blanche. L'analyse voxel à voxel a permis d'identifier une zone située à l'intersection des voies dorsale et ventrale du langage, au-dessous de la jonction temporo-pariétale (JTP) gauche. La gravité des dommages dans cette région, évaluée par les valeurs de la FA, était mieux corrélée aux déficits phasiques que le volume de l'infarctus. Dans cette thèse, nous avons trouvé que, après un AIC, l'atteinte du FCS est un bon prédicteur de la sévérité du déficit moteur et celle d'une région localisée au-dessous de la JTP gauche, est un bon prédicteur de la sévérité de l'aphasie. / The diffusion tensor imaging (DTI) is a technique used to study the organization and the structural state of white matter tracts. The study of the tracts integrity may help to understand and determine the severity of the patients? prognosis after an ischemic stroke. In this thesis, I present two studies: in the first we compared the results obtained from a template of the Corticospinal tract (CST) to those obtained by a tractography in healthy subjects and stroke patients. In both groups, the fractional anisotropy (FA) values obtained with the tractography were higher than those obtained by the template. However, in patients, both methods detected a decrease in FA values of the ipsilesional CST, which was correlated with motor scores, but the FA values obtained with the template were better correlated with the BOLD signal generated by the paretic hand. The second study investigated whether the severity of aphasia could be correlated to the damage of certain areas of the white matter. The voxel-based analysis identified an area at the intersection of the dorsal and ventral pathways of language, below the left temporo-parietal junction (JTP). The severity of the damage in this area, assessed by the FA values, was better correlated with the phasic deficits than the infarct volume. In this thesis, we found that after a stroke, the damage of the CST is a good predictor of the motor deficit severity and that of the region located below the left JTP is a good predictor of the aphasia severity.

Diploma thesis focus on the regulation of consumer housing loans market. Based on the analysis of the consumer housing loans market in the European Un-ion, the main regulations and factors affecting this market are identified. Panel data analysis is available to determine the effects of regulation and factors on mortgage lending. The explained variable is the growth of mortgage loans. The explanatory variables are macroeconomic factors and macroprudential policy in-struments. Results showing effects of macro-prudential instruments on the con-sumer housing loans market.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第19270号 / 医博第4034号 / 新制||医||1011(附属図書館) / 32272 / 京都大学大学院医学研究科医学専攻 / (主査)教授 髙橋 良輔, 教授 渡邉 大, 教授 富樫 かおり / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Introduction: Somatosensory evoked potentials (SSEP) play a pivotal role in the diagnosis and disease monitoring of multiple sclerosis (MS). Delayed latencies are a surrogate for demyelination along the sensory aerence. This study aimed to evaluate if SSEP latencies are representative of demyelination of the brain overall, by correlating with cerebral microstructural integrity as measured by Magnetic resonance (MR) diusion tensor imaging (DTI). Analysis was performed in a hypothesis-free whole brain approach using tract-based spatial statistics (TBSS). Material and methods: A total of 46 patients with MS or clinically isolated syndrome were included in the study. Bilateral SSEPs of the median nerve measuring mean N20 latencies (mN20) and Central Conduction Time (CCT), were acquired. MRI scans were performed at 3T. DTI acquisition was done with a single-shot echoplanar imaging technique with 80 diusion directions. The FSL software package was used to process the DTI datasets and to calculate maps of fractional anisotropy (FA), axial diusivity (AD), and radial diusivity (RD). These maps were then further analyzed using the TBSS module. The mean N20 and CCT and the right- and left-sided N20 and CCT were separately correlated to FA, AD, and RD, controlled for age, gender, and EDSS as variables of non-interest. Results: Widespread negative correlations of SSEP latencies with FA (p = 0.0005) and positive correlations with RD (p = 0.0003) were measured in distinct white matter tracts, especially the optic tracts, corpus callosum, and posterior corona radiata. No correlation with AD was found in any white matter tract. Conclusion: Highly significant correlations of FA and RD to SSEPs suggest that their latency is representative of widespread microstructural change, and especially demyelination in patients suering from MS, reaching beyond the classic somatosensory regions. This points to the usefulness of SSEPs as a non-invasive tool in the evaluation of microstructural damage to the brain.

No description available.