Research on the Segmentation of MRI Image Based on Immune Support Vector Machine

In head MRI image, the boundary of each encephalic tissue is highly complicated and irregular. It is a real challenge to traditional segmentation algorithms. Support vector machine (SVM) has high generalization ability, especially for dataset with small number of samples in high dimensional space. However, selecting parameters for SVM is a complicated problem which directly affects segmentation result. immune algorithm (IA), mainly applied to optimization, has the abilities of learning, memorizing and self- adaptive adjusting. The main idea is to search optimal parameters for SVM using IA. In this paper, an immune support vector machine (ISVM) is proposed to segment MRI image. As our experiment shown, the boundaries of 5 kinds of encephalic tissues are extracted successfully, and ISVM reaches satisfactory generalization accuracy.