A Modified Distance Regularized Level Set Evolution for Masseter Segmentation

Abstract

Segmentation of the masseter muscle is one of the most important bases for diagnosis and treatment, but automatical segmentation of masseter is very difficult as the masseter and its adjacent tissues have very similar gray levels. In this paper, a novel method is presented to segment the masseter in MRI images, which modifies the distance regularized level set evolution (DRLSE) with a new adaptive edge indicator function. The presented method introduces the phase congruency into the edge indicator function and combines the image gradient with phase information together to solve the segmentation problem. The method is tested by 50 MR images of masseter muscle, and the results show it is an effective approach can be used to produce clinically acceptable results to this challenging work.