Detecting Anatomical Changes Using Logistic Regression of Structure Masks

Abstract

Listen

Translate article

Introduction Neuroscientists are often interested in determining whether external covariates are related to changes in brain shape. There are three main ways currently being used to determine anatomical changes in human brains: 1) deformation fields produced from nonlinear warping techniques [1-3]; 2) surface extraction techniques [4,6,7,10] and 3) voxel analysis of structure masks [9]. Using results from random field theory, we show how logistic regression of structure masks can be used to detect anatomical changes.