Performance Evaluation of Segmentation Algorithms Based on Level Set Method: Application to Medical Images

Currently, the phase of segmentation is an important step in the treatment and the interpretation of the medical images; it represents one of the most difficult step for the extraction of the relevant parameters of the image and fact part of a very active field and rich of research. In this paper, we present an overview about segmentation methods based on level set technique, namely Caselle method, Chan Vese method, Chumming Li method, Lankton method, Bernard method and Shi method. The performance of each method can be evaluated either visually, or from similarity measurements between a reference and the results of the segmentation. We have applied each method for different medical images. We present a comparative evaluation of the considered segmentation methods, with respect to four criteria, given specific medical datasets. Through simulated results, we have demonstrated that the best results are achieved by Shi method and Chan & Vese method.