A Multiscale Methodology for Mapping Human Neuroanatomy

This chapter describes how to use Mode-based Morphometry (MBM) as a new approach to map anatomical variations in the human brain at multiple spatial scales. This approach overcomes the limitations of conventional approaches, such as surface-based morphometry or voxel-based morphometry, which only consider localized and independent effects at specific brain locations. In contrast, MBM analyzes anatomical group differences of a brain structure as a linear combination of its geometric eigenmodes, which represent intrinsic resonant patterns spanning multiple spatial frequencies that are determined by the geometry of the cortex. Assessing the contributions that each mode makes to the pattern of anatomical differences sheds light on the spatial scales at which such differences are preferentially expressed. MBM can reveal the underlying spatial pattern of group differences, enhance reproducibility, and offer insights into the generative mechanisms that shape the observed differences. Here, we provide step-by-step guideline for conducting an MBM analysis with example data. The approach can be applied to a wide range of anatomical data, given a geometric surface or volume of the structure to derive the geometric eigenmodes.