| Mouse models of Huntington's disease (HD) play an important role in investigating the disease mechanisms and testing potential therapeutic treatments. Progressive atrophy of the brain, particularly the striatum, is a characteristic phenotype of HD, and is known to begin long before the onset of motor symptoms. Elucidating the spatial and temporal patterns of atrophy in HD mouse models is therefore important to characterize the phenotypes of these models, as well as evaluate the effects of neuroprotective treatments at specific time frames during disease progression. The objective of this thesis was to develop a combined approach based on in vivo MRI, creation of a population-averaged reference brain atlas, and deformation-based longitudinal morphological analysis, in order to investigate the rate and spatiotemporal progression of brain atrophy in longitudinal studies of mouse models of HD.;Based on a statistical framework using mixed-effects modeling of deformation-based metrics, the proposed techniques provided an unbiased, exploratory mapping of age-related morphological changes in the wild-type mouse brains, and pathology-induced progressive brain atrophy in the R6/2 and N171-82Q mouse models of HD. The longitudinal morphological analysis framework was also used to evaluate and grade the effects of sertraline and coenzyme Q10 (CoQ10) treatments on the rate of progressive atrophy in the N171-82Q model of HD. Further, progressive cortical and striatal atrophy in the N171-82Q HD mice showed significant positive correlations with measured functional deficits in this model, suggesting that MRI-based metrics can be used as reliable complementary markers for longitudinal assessment of disease severity in HD mouse models. The findings of this thesis provide proof-of-principle for the application of in vivo MRI based morphological analysis for automated mapping of structural brain atrophy in longitudinal studies of HD mouse models, and will be useful for future testing and comparison of potential therapeutics in these models. |