Amygdalar shape analysis method using surface contour aligning, spherical mapping, and probabilistic subregional segmentation

Abstract

The objective of this study was to develop a reliable method for the shape analysis of the amygdala, a structure that is important in gaining a better understanding of the limbic system in the human brain. The goal of this study was threefold; to develop (1) a robust method for aligning the contour of the amygdala; (2) a reproducible method for extracting surface parameters of the amygdala using a spherical mapping technique; and (3) a standardized approach for statistical assessment and visualization of shape alterations by applying the probabilistic maps of amygdalar subregions. This technique was validated by conducting an artificial phantom study and by assessing sex-related amygdalar shape differences using T1-weighted images from healthy volunteers. In the phantom study, the region with atrophy was detected successfully through the shape analysis process. In the human study, the average radii of the centromedial (CM) subregion in the left amygdala and laterobasal (LB), superficial (SF) and CM subregions in the right amygdala were different between sexes (t-tests. p=0.02, 0.04, 0.04, and 0.002, respectively). In addition, focal regions with larger radii in amygdalae of men than those of women were found predominantly on the surfaces of bilateral SF and bilateral CM subregions, after the volumes of the amygdala had been scaled to the unit volume (1000 mm(3)) (Mann-Whitney U-test, false discovery rate corrected p<0.05, clustered vertex points >25). Regions with smaller radii in amygdalae of men were found predominantly on the anterior surfaces of the right LB and SF subregions (Mann-Whitney U-test, false discovery rate corrected p<0.05, clustered vertex points >25). This is generally in agreement with previous findings from animal studies. The current method may be used for measuring subtle local shape changes of the amygdala in various psychiatric or neurologic disorders. (C) 2010 Elsevier Ireland Ltd. All rights reserved.