We will develop and apply novel geometric algorithms to brain diffusion MRI images obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and Australian twin datasets. Despite major advances in brain imaging measures on grey matter, there is still a lack of efficient brain multimodal imaging algorithms by integrating brain white matter and grey matter morphology. In the current project, we will develop novel diffusion MRI analysis methods using quasiconformal geometry and volumetric harmonic map. Specifically, we will develop 1) multimodal brain surface parcellation for tractography and network analysis, and 2) brain white matter multivariate tensor-based morphometry to measure white matter morphometry. The proposed system will focus on developing a coherent computational framework to integrate both white and grey matter morphology and compute their complete geometric features to capture subtle brain white matter changes. This may provide an effective way to pinpoint unique brain network structure or subregional areas for diagnostic group difference comparison and identify genetic effects that might account for some of the structural variations in a large population. To investigate the reliability and practicality of our method, we seek to 1) construct structural brain network with tractography on multimodal surface parcellation results, and 2) compute brain white matter multivariate tensor-based morphometry (WM-mTBM) features. The computed connectivity network measures and the WM-mTBM features will be validated by 1) between-diagnostic-group comparisons in the ADNI dataset, and 2) genetic influence analysis in the Australian twin dataset.
This project proposed a novel brain imaging system which integrates brain white and grey matter morphology. It will be applied for clinical assessment of mild cognitive impairment (MCI) and Alzheimer's disease (AD), and for understanding how genetic factors affect brain connectivity.
Showing the most recent 10 out of 25 publications
