Hippocampal morphology and cytoarchitecture in the 3D BigBrain

Abstract

Abstract The internal architecture of the hippocampus is challenging to map in detail using traditional histology and in-vivo neuroimaging. This is due, in part, to its complex archicortical folding that is difficult to appreciate in both modalities. Here, we aimed to overcome this challenge by leveraging the unique histological dataset available as open-source 3D BigBrain. Specifically, we investigated the relationship between topology, laminar cytoarchitecture, and detailed morphology with respect to hippocampal subfields and its anterior-posterior axis. Inspired by computational parcellation methods used in the neocortex, we topologically ‘unfolded’ the hippocampus and mapped it with respect to 5 morphological and 10 laminar features. Several features, including thickness, gyrification, and mean neuronal density, clearly differed between subfields. Indeed, data-driven clustering of all features revealed subdivisions which closely resemble manually defined subfields. Some features, most notably gyrification, also showed anterior-posterior differences within subfields, which may relate to connectivity and functional differences described in previous literature. Overall these findings offer quantifiable markers of hippocampal subfields, and provide new anatomical insight into the topology and properties of hippocampal tissue. Future applications could involve translation to in-vivo MRI for probing the internal hippocampal architecture at this mesoscale in cognition and disease.