Automatic shape asymmetry analysis of human cerebral hemispheres in schizophrenia based on MRI

Abstract

Magnetic Resonance Imaging (MRI) based brain asymmetry studies have reported that the magnitude and direction of the structural asymmetry of human brain could be affected by gender, handedness and mental diseases. The studies always required manual interaction in the extraction of the brain volumes. This thesis presents a novel fully automatic method based on 3D MRI to analyze the difference of shape asymmetry between two populations. The main focus of this thesis is the asymmetry analysis. The method consists of four major steps: segmentation of cerebral hemispheres in MRI, alignment of cerebrum volumes into the stereotaxic space with affine registration, the extraction of cerebral hemisphere surfaces, and finally statistical analysis of the cerebral shape asymmetry. The calculated differences of the estimated shape asymmetries between the two evaluated populations showed significant similarities in the shape asymmetries. /Kir09