Automatic Segmentation of Brain MRI Images by Threshold Level Set Method

Abstract

A threshold level set method with adaptive parameters applied on automatic segmentation of brain MRI image is presented in this paper. Threshold level set method is one kind of the active contour models applied on the image segmentation. There are two parameters of threshold level set for adjusting threshold purposes. These two thresholds adjust the speed function for controlling the stop position of the contours. The optimal parameters are automatically decided by using the adaptive parameter selection algorithm. The contours of brain white matter are segmented with MRI images by both the developed method and manual operation. By experiment results, it is demonstrated that the proposed method is effective and efficient than manual selection of parameters.