My research portfolio combines theoretical work on statistical shape characterization using a symmetry-based representation of shape with applied and translational research in multiple areas of biomedical image analysis. I am particularly interested in analysis techniques that are tailored to specific anatomical structures. My key work in this area involves automatic segmentation and morphometry of the hippocampal formation (HF) in magnetic resonance imaging (MRI). The HF plays a central role in memory function and is a site of early neurodegeneration in Alzheimer’s disease. In a broad effort to develop more detailed imaging-based biomarkers for Alzheimer’s disease, my team has developed a first of its kind detailed computational atlas of the HF from postmortem MRI microscopy. With NIH R01 funding, we are now integrating this atlas with histology, leveraging it for the analysis of the HF and its subfields in in vivo MRI, and evaluating MRI-derived subfield-specific as alternative imaging biomarkers in dementia. We recently published a pioneering technique for automatic segmentation of HF subfields in high-resolution, HF-focused T2-weighted in vivo MRI, showing excellent agreement with manual segmentation. Concurrently, we developed and evaluated a strategy for improving the accuracy of general-purpose segmentation techniques using machine learning as a corrective wrapper method. Other recent publications include a surface-based tract-specific strategy for diffusion MRI analysis, for which open-source software is currently being developed, and an automatic cardiac MRI segmentation method with an explicit prior on heart wall thickness.