Abstract:
The main goal of this thesis is to find distinct macro-structural characteristics of brain white matter in the case of psychosis, where development of diagnostic imaging measures is necessary for early diagnosis and prospective studies. Given a tractogram data, which is a dense set of white matter fiber pathways of the whole brain obtained from diffusion magnetic resonance imaging, we propose to compute a global measure of dispersion for a voxel from the end point statistics of a set of fibers, which indicates complexity of the white matter voxel not locally but at macro scales. The findings on phantom data demonstrate sensitivity of the proposed measure to the tuning parameters and show its range characteristics. The findings on the real data demonstrate that proposed macro-structural dispersion information is found to be significant for discrimination of the schizophrenia and the bipolar patients from the healthy controls, especially when the frontally associative bundles such as cingulum and inferior occipito-frontal fasciculus are considered. The macroscopic dispersion measure is as informative as the local diffusion measures for the detection of changes in the white matter regions due to the psychosis. Beside, as a technical application, the dispersion map is considered and experimented for segmentation of cingulum. The findings of the thesis provide that the proposed measure is a potential diagnostic imaging marker in the case of psychosis and we contribute to the field of diagnostic research by generating a novel dispersion map of the brain that could be used for other clinical and technical applications.|Keywords : Brain white matter, macroscopic dispersion, tractogram, bipolar, schizophrenia.
