Shape Analysis in Medical Imaging

形状分析是医学成像的基础。本研究旨在探索和利用几何形状的作用，在制定基于形状的生物标志物的疾病影响脑形态。具体目标是（i）开发基于形状的模型，以量化大脑结构中几何和拓扑结构的多尺度变化，（ii）开发统计框架，将联合收割机异构数据集与未标记的神经成像数据相结合，以及（iii）确定脑萎缩中的形状动力学，以识别神经退行性变的风险。形状分析的这些进展将提供一个新的基于形状的生物标志物家族，使用大量未开发的医学成像数据发现解剖结构的更小细节。

Shape analysis is fundamental to medical imaging. This research aims to explore and exploit the role of geometry in the elaboration of shape-based biomarkers for diseases affecting brain morphology. The specific objectives are to (i) develop shape-based models to quantify multiscale variations of geometry and topology in brain structures, (ii) develop statistical frameworks to combine heterogeneous datasets with unlabeled neuroimaging data, and (iii) determine shape dynamics in brain atrophy to identify a risk of neurodegeneration. These advances in shape analysis will provide a new family of shape-based biomarkers, discovering much smaller details of anatomy using largely unexplored medical imaging data.