Principled clustering of functional signals reveals gradients in processing both within the anterior hippocampus and across its long axis

Abstract

AbstractA particularly elusive puzzle concerning the hippocampus is how the structural differences along its long, anteroposterior axis might beget meaningful functional differences, particularly in terms of the granularity of information processing. One measure posits to quantify this granularity by calculating the average statistical independence of the BOLD signal across neighboring voxels, or inter-voxel similarity (IVS), and has shown the anterior hippocampus to process coarser-grained information than that in the posterior hippocampus. This model of the hippocampus, however, conflicts with a number of task-oriented findings, many of which have varied in their fMRI acquisition parameters and hippocampal parcellation methods. In order to reconcile these findings, we measured IVS across two separate resting-state fMRI acquisitions and compared the results across many of the most widely used parcellation methods in a large young-adult sample (Acquisition 1, N = 253; Acquisition 2, N = 183). Finding conflicting results across acquisitions and parcellations, we reasoned that a principled, data-driven approach to hippocampal parcellation is necessary. To this end, we implemented a group masked independent components analysis (mICA) to identify functional subunits of the hippocampus, most notably separating the anterior hippocampus into separate anterior-medial, anterior-lateral, and posteroanterior-lateral components. Measuring IVS across these components revealed a decrease in IVS along the medial-lateral axis of the anterior hippocampus but an increase from anterior to posterior. We conclude that representational granularity may not change linearly or unidirectionally across the hippocampus, and that moving the study of the hippocampus towards reproducibility requires grounding it in a functionally informed approach.Significance StatementProcessing information along hierarchical scales of granularity is critical for many of the feats of cognition considered most human. Recently, the changes in structure, cortical connectivity, and apparent functional properties across parcels of the hippocampal long axis have been hypothesized to underlie this hierarchical gradient in information processing. We show here, however, that the choice of parcellation method itself drastically affects the perceived granularity across the hippocampus, and that a principled, functionally informed approach to parcellation reveals gradients both within the anterior hippocampus and in non-linear form across the long axis. These results point to the issue of parcellation as a critical one in the study of the hippocampus and reorient interpretation of existing results.