J. Martin, A. Pentland, R. Kikinis

Abstract

This paper presents a framework for analyzing the shape deformation of structures within the human brain. A mathematical model is developed describing the deformation of any brain structure whose shape is a ected by both gross and detailed physical processes. By using the equivalence of physically-based and probability-based formulations of deformation, the total shape deformation is decomposed into physical modes of variation obtained from nite element analysis, and experimental modes of variation obtained from sample data using principal component analysis. This mathematical model is used to classify diseases that a ect the shape of the ventricular system of the brain. Because ventricular shape is a ected not only by pathology but also by overal l brain shape, disease discrimination is dicult. By modeling the brain's elastic properties, we are able to compensate for some of the nonpathological modes of ventricular shape variation. This al lows us to experimental ly characterize modes of variation that are indicative of disease processes. Our technique is applied to magnetic resonance images of the brains of individuals with Alzheimer's disease and normal-pressure hydrocephalus, as wel l as to healthy volunteers. Classi cation results are presented.