Automated segmentation of brain exterior in MR images driven by empirical procedures and anatomical knowledge

Abstract

This work demonstrates encouraging initial results for increasing the automation of a practical and precise MR brain image segmentation method. The intensity threshold for segmenting the brain exterior is automatically determined by locating the choroid plexus. This is done by finding peaks in a series of histograms taken over regions specified using anatomical knowledge. Intensity inhomogeneities are accounted for by adjusting the global intensity to match the white matter peak intensity in local regions. The results from 20 different brain scans (over 1000 images) obtained under different conditions are presented to validate the method which was able to determine the appropriate threshold in approximately 80% of the data.