During my time at the Martinos Center, we developed a computational atlas of the hippocampus at the substructure level using ultra-high resolution ex vivo MRI data. This atlas can be used in a Bayesian segmentation framework in order to automatically label the hippocampal substrcutures in in vivo MRI brain scans. The generative nature of the approach enables the segmentation method to automatically adapt to the MRI contrast of the input, e.g., T1-weighted or T2-weighted scans. Moreover, it can seamlessly analyze multimodal MRI data, as in the example above.

An implementation of the segmentation method can be found in the popular neuroimaging package FreeSurfer.

Instructions on how to download and use the software (cross-sectional version)

A related post on the potential applications of hippocampal subfield analysis    (versión en Español)