Assessing structural alterations in spinocerebellar ataxias using fractal dimension analysis and modular analysis from magnetic resonance brain images

• Main Authors: Tzu-Yun Wang, 王姿勻
• Other Authors: Yu-Te Wu
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/16860461227016709500

博士 === 國立陽明大學 === 生物醫學影像暨放射科學系 === 103 === Background: Ataxic disorders are neurodegenerative disease that primarily affect cerebellum, which are characterized as progressive ataxia and a variable pattern of other neurological deficits. The most common forms of ataxic disorders are spinocerebellar ataxias (SCAs), such as SCA1, SCA2 and SCA3, and the multiple system atrophy cerebellar type (MSA-C) also shows clinical symptoms similar to SCAs. We hypothesize that the clinical features of neurodegenerative ataxia, such as scale for the assessment and rating of ataxia (SARA), cortical morphology, may be specifically related to certain combination of different types of gene expansion. Moreover, the changes of cortical morphology in different functional regions can be altered due to ataxic disorders. Methods: In this study, the three-dimensional fractal dimension (3D-FD) method was applied to quantify the changes of cortical morphology in ataxic patients. We attempt to investigate the supratentorial involvement of ataxic patients by measuring significant atrophy in relatively focal regions. Since we observed that joint distributions over the atrophic brain regions were often dependent and the marginal distributions in atrophic brain regions maybe not completely normal distribution, the use of Pearson correlation to calculate the inter-regional connectivities can be problematic. Instead, we used the copula correlation to measure the correlation, which is similar to nonparametric rank-based correlation coefficient estimators, and has the advantage in preventing the bias based on asymmetric distributions. Based on the interregional connectivity, we utilized copula-derived modular analysis to categorize the altered regions of ataxic patients into several major functional communities. By comparing the structural connectivity pattern between ataxic patients and healthy controls, the variation of copula-derived modular network may allow us to measure the inherently mental state of ataxic patients. Furthermore, we used the mutual information (MI) to analyze validity between the phenotype of gene in ataxic patients and the altered regions, and to further assess the association between expressions of two genes. Results: Forty-eight genetically confirmed SCA3 patients, one hundred and thirteen MSA-C patients and fifty gender- and age-matched control persons were recruited in this study. Using the 3D-FD method, we found that the cortical involvement in SCA3 was more extensive than involvement of traditional olivopontocerebellar regions and the corticocerebellar system. Moreover, the significant correlation between decreased 3D-FD values and disease duration may indicate atrophy of the cerebellar cortex and cerebral cortex in both SCA3 and MSA-C patients. Additionally, the modular network of SCA3 exhibited primary three modules, which corresponded to movement-related, cognitive-related, visual-spatial-related and emotional/mnemonic-related functional regions, respectively. In this thesis, our results indicated that the cognitive decline and visual-spatial related defects of SCA3 may result from the implication of disrupted cortico-cerebellar circuit, which involves the non-motor functions as well as motor networks in the pathophysiology of SCA3. Eventually, we found that both in SCA3 and MSA-C patients, their CAG repeat numbers of SCA1~3, SCA6~8, SCA10, SCA12, SCA17 and dentatorubral-pallidoluysian atrophy (DRPLA) were relevant to each other. This result suggests that there may be the informational interaction between different phenotype of gene in both SCA3 and MSA-C patients.