Distinct Structural Patterns of the Human Brain: A Caveat for Registration

Abstract

Current approaches for analyzing structural patterns of the human brain often implicitly assume that brains are variants of a single type, and use nonlinear registration to reduce the inter-individual variability. This assumption is challenged here. Regional anatomical and connection patterns cluster into statistically distinct types. An advanced analysis proposed here leads to a deeper understanding of the governing principles of cortical variability.