京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第18854号 / 医博第3965号 / 新制||医||1007(附属図書館) / 31805 / 京都大学大学院医学研究科医学専攻 / (主査)教授 古川 壽亮, 教授 髙橋 良輔, 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

京都大学 / 0048 / 新制・論文博士 / 博士(医学) / 乙第13120号 / 論医博第2133号 / 新制||医||1023(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 高橋 淳, 教授 村井 俊哉, 教授 林 康紀 / 学位規則第4条第2項該当 / Doctor of Medical Science / Kyoto University / DFAM

Theory of Mind (ToM) is the ability to infer mental states of others and this skill relies on a distributed network of brain regions. A brain region that has been traditionally disregarded in relation to non-motor functions is the cerebellum. Here, we leveraged large-scale multimodal neuroimaging data to elucidate the structural and functional role of the cerebellum in ToM. We used functional activations to determine whether the cerebellum has a domain-general or domain-specific functional role. We found that the cerebellum is organized in a domain-specific way. We used effective connectivity and probabilistic tractography to map the cerebello-cerebral ToM network. We found a left cerebellar effective and structural lateralization, with more and stronger effective connections from the left cerebellar hemisphere to the contralateral cerebral ToM areas and greater cerebello-thalamo-cortical (CTC) and cortico-ponto-cerebellar (CPC) streamline counts from and to the left cerebellum. Lastly, we examined the relationship between CTC and CPC white matter and ToM speed and accuracy but found no correlation. Our study provides novel insights to the network organization of the cerebellum, an overlooked brain structure, and ToM, one of humans’ most essential abilities to navigate the social world. / Psychology

Recent advances in tractography allow for connectomes to be constructed in vivo. These have applications for example in brain tumor surgery and understanding of brain development and diseases. The large size of the data produced by these methods lead to a variety problems, including how to evaluate tractography outputs, development of faster processing algorithms for tractography and clustering, and the development of advanced visualization methods for verification and exploration. This thesis presents several advances in these fields. First, an evaluation is presented for the robustness to noise of multiple commonly used tractography algorithms. It employs a Monte–Carlo simulation of measurement noise on a constructed ground truth dataset. As a result of this evaluation, evidence for obustness of global tractography is found, and algorithmic sources of uncertainty are identified. The second contribution is a fast clustering algorithm for tractography data based on k–means and vector fields for representing the flow of each cluster. It is demonstrated that this algorithm can handle large tractography datasets due to its linear time and memory complexity, and that it can effectively integrate interrupted fibers that would be rejected as outliers by other algorithms. Furthermore, a visualization for the exploration of structural connectomes is presented. It uses illustrative rendering techniques for efficient presentation of connecting fiber bundles in context in anatomical space. Visual hints are employed to improve the perception of spatial relations. Finally, a visualization method with application to exploration and verification of probabilistic tractography is presented, which improves on the previously presented Fiber Stippling technique. It is demonstrated that the method is able to show multiple overlapping tracts in context, and correctly present crossing fiber configurations.

Background: Tractography has become a standard technique for planning neurosurgical operations in the past decades. This technique relies on diffusion magnetic resonance imaging. The cutoff value for the fractional anisotropy (FA) has an important role in avoiding false-positive and false-negative results. However, there is a wide variation in FA cutoff values. Methods: We analyzed a prospective cohort of 14 patients (six males and eight females, 50.1 ± 4.0 years old) with intracerebral tumors that were mostly gliomas. Magnetic resonance imaging (MRI) was obtained within 7 days before and within 7 days after surgery with T1 and diffusion tensor image (DTI) sequences. We, then, reconstructed the corticospinal tract (CST) in all patients and extracted the FA values within the resulting volume. Results: The mean FA in all CSTs was 0.4406 ± 0.0003 with the fifth percentile at 0.1454. FA values in right-hemispheric CSTs were lower (p < 0.0001). Postoperatively, the FA values were more condensed around their mean (p < 0.0001). The analysis of infiltrated or compressed CSTs revealed a lower fifth percentile (0.1407 ± 0.0109 versus 0.1763 ± 0.0040, p = 0.0036). Conclusion: An FA cutoff value of 0.15 appears to be reasonable for neurosurgical patients and may shorten the tractography workflow. However, infiltrated fiber bundles must trigger vigilance and may require lower cutoffs.

No description available.

No description available.

Tractography is a tool that is often used to study structural brain connectivity from diffusion magnetic resonance imaging data. Despite its ability to visualize fibers in the white brain matter, it results in a high number of invalid streamlines. For the sake of research and clinical applications, it is of great interest to reduce this number and so improve the quality of tractography. Over the years, many solutions have been proposed, often with a need for ground truth data. As such data for tractography is very difficult to generate even with expertise, it is meaningful to instead use methods like reinforcement learning that does not have such a requirement. In 2021 a deep reinforcement learning tractography network was published: Track-To-Learn. There is however still room for improvement in the reward function of the framework and this is what we focused on in this thesis. Firstly we successfully reproduced some of the published results of Track-To-Learn and observed that almost 20 % of the streamlines were anatomically plausible. Continuously we modified the reward function by giving a reward boost to streamlines which started or terminated within a specified mask. This addition resulted in a small increase of plausible streamlines for a more realistic dataset. Lastly we attempted to include anatomical filtering in the reward function. The produced results were however not enough to draw any valid conclusions about the influence of the modification. Nonetheless, the work of this thesis showed that including further fiber specific anatomical constraints in the reward function of Track-To-Learn could possibly improve the quality of the generated tractograms and would be of interest in both research and clinical settings.

Diffusion Tensor Magnetic Resonance Imaging (DT-MRI), also known as Diffusion Tensor Imaging (DTI), is a unique medical imaging modality that provides non-invasive estimates of White Matter (WM) connectivity based on local principal directions of anisotropic water diffusion. DTI tractography estimates are a macroscopically sampled description of underlying microscopic structure, and are therefore of limited validity. The under-sampling of underlying white matter structure in DTI data gives rise to Intra-Voxel Orientational Heterogeneity (IVOH), a condition in which white matter structures of multiple different orientations are averaged into a single DTI voxel sample, causing a loss of validity in the diffusion tensor model. Fast Marching Tractography (FMT) algorithms based on fast marching level set methods have been proposed to better handle the presence of IVOH in DTI data when compared to older Streamline Tractography (SLT) methods. However, the actual performance advantage of any tractography algorithm over another cannot be conclusively stated until a ground truth standard of comparison is developed. This work develops an optimized version of the FMT algorithm that is dubbed the Front Propagation Tractography (FPT) algorithm. The FPT algorithm includes unique approaches to the speed function, connectivity estimation, and likelihood estimation components of the FMT framework. The performance of the FPT algorithm is compared against the SLT algorithm using ground truth software phantom data and human brain data. Software phantom ground truth experiments compare the performance of each algorithm in single tract and crossing tract structures for varying levels of diffusion tensor field perturbation. Human brain estimates in the corpus callosum yield qualitative comparisons from inspection of 3D visualizations. A final area of exploration is the construction and analysis of a ground truth physical DTI phantom manifesting IVOH. / Master of Science

La structure et l'organisation de la substance blanche du cerveau humain ne sont pas encore complètement connues. L'Imagerie par Résonance Magnétique de diffusion (IRMd) offre une approche unique pour étudier in vivo la structure des tissus cérébraux, permettant la reconstruction non invasive des trajectoires des faisceaux de fibres du cerveau en utilisant la tractographie. Aujourd'hui, les techniques récentes d'IRMd avec haute résolution angulaire (HARDI) ont largement amélioré la qualité de la tractographie par rapport à l'imagerie du tenseur de diffusion standard (DTI). Toutefois, les jeux de données de tractographie résultant sont très complexes et comprennent des millions de fibres, ce qui nécessite une nouvelle génération de méthodes d'analyse. Au-delà de la cartographie des principales voies de la substance blanche, cette nouvelle technologie ouvre la voie à l'étude des faisceaux d'association courts, qui ont rarement été étudiés avant et qui sont au centre de cette thèse. L'objectif est d'inférer un atlas des faisceaux de fibres du cerveau humain et une méthode qui permet le mappage de cet atlas à tout nouveau cerveau.Afin de surmonter la limitation induite par la taille et la complexité des jeux de données de tractographie, nous proposons une stratégie à deux niveaux, qui enchaîne des regroupements de fibres intra- et inter-sujet. Le premier niveau, un regroupement intra-sujet, est composé par plusieurs étapes qui effectuent un regroupement hiérarchique et robuste des fibres issues de la tractographie, pouvant traiter des jeux de données contenant des millions de fibres. Le résultat final est un ensemble de quelques milliers de faisceaux de fibres homogènes représentant la structure du jeu de données de tractographie dans sa totalité. Cette représentation simplifiée de la substance blanche peut être utilisée par plusieurs études sur la structure des faisceaux individuels ou des analyses de groupe. La robustesse et le coût de l'extensibilité de la méthode sont vérifiés à l'aide de jeux de fibres simulés. Le deuxième niveau, un regroupement inter-sujet, rassemble les faisceaux obtenus dans le premier niveau pour une population de sujets et effectue un regroupement après normalisation spatiale. Il produit en sortie un modèle composé d'une liste de faisceaux de fibres génériques qui peuvent être détectés dans la plupart de la population. Une validation avec des jeux de données simulées est appliqué afin d'étudier le comportement du regroupement inter-sujet sur une population de sujets alignés avec une transformation affine. La méthode a été appliquée aux jeux de fibres calculés à partir des données HARDI de douze cerveaux adultes. Un nouveau atlas des faisceaux HARDI multi-sujet, qui représente la variabilité de la forme et la position des faisceaux à travers les sujets, a été ainsi inféré. L'atlas comprend 36 faisceaux de la substance blanche profonde, dont certains représentent quelques subdivisions des faisceaux connus, et 94 faisceaux d'association courts de la substance blanche superficielle. Enfin, nous proposons une méthode de segmentation automatique de mappage de cet atlas à tout nouveau sujet. / Human brain white matter (WM) structure and organisation are not yet completely known. Diffusion-Weighted Magnetic Resonance Imaging (dMRI) offers a unique approach to study in vivo the structure of brain tissues, allowing the non invasive reconstruction of brain fiber bundle trajectories using tractography. Nowadays, the recent dMRI techniques with high angular resolution (HARDI) have largely improve the quality of tractography relative to standard diffusion tensor imaging. However, the resulting tractography datasets are highly complex and include millions of fibers which requires a new generation of analysis methods. Beyond the mapping of the main white matter pathways, this new technology opens the road to the study of short association bundles, which have been rarely studied before and is in the focus of this thesis. The goal is to infer an atlas of the fiber bundles of the human brain and a method mapping this atlas to any new brain.In order to overcome the limitation induced by the size and complexity of the tractography datasets, we propose a two-level strategy, chaining intra- and inter-subject fiber clustering. The first level, an intra-subject clustering, is composed by several steps performing a robust hierarchical clustering of a fiber tractography dataset that can deal with millions of diffusion-based tracts. The end result is a set of a few thousand homogeneous bundles representing the whole structure of the tractography dataset. This simplified representation of white matter can be used further for several studies of individual bundle structure or group analyses. The robustness and the cost of the scalability of the method are checked using simulated tract datasets. The second level, an inter-subject clustering, gathers the bundles obtained in the first level for a population of subjects and performs a clustering after spatial normalization. It produces as output a model composed by a list of generic fiber bundles that can be detected in most of the population. A validation with simulated datasets is applied in order to study the behavior of the inter-subject clustering over a population of subjects aligned with affine registration. The whole method was applied to the tracts computed from HARDI data obtained for twelve adult brains. A novel HARDI multi-subject bundle atlas, representing the variability of the bundle shape and position across subjects was thus inferred. The atlas includes 36 deep WM bundles, some of these representing a few subdivisions of known WM tracts, and 94 short association bundles of superficial WM. Finally, we propose an automatic segmentation method mapping this atlas to any new subject.