Precon_all: A species-agnostic automated pipeline for non-human cortical surface reconstruction

Abstract

Cortical surface reconstruction has transformed the study of brain morphology and cortical geometry in human neuroimaging. Here we present precon_all, an open-source structural imaging pipeline that generalizes reconstruction of the neocortical surface to multiple species. Adopting a modular framework that mirrors FreeSurfer's recon-all in humans, precon_all produces surfaces for non-human species compatible with FreeSurfer and Connectome Workbench. Precon_all reduces the dependencies of surface reconstruction to an image with T1-like contrast and five user-defined masks. This removes the dependence on species-specific atlases and tissue priors, and in turn extends surface reconstruction to species where these resources do not yet exist. The pipeline supports a wide range of imaging conditions, including in-vivo and post-mortem data, and has thus far been used to reconstruct the cortical geometry of primates, carnivores, and ungulates across multiple sites, spanning both highly gyrified and lissencephalic brains. Precon_all can be automated for high-throughput processing, and its template construction module can build surface-based common spaces to enable group-level analysis of cortical morphometry in new species. In short, precon_all is a flexible, modular pipeline that enables the comparative neuroimaging community to study brain evolution through cortical geometry in an ever-growing number of species.