Technology Research, Medical Image Segmentation Based On Fuzzy Clustering

Medical image segmentation has been playing an increasingly important role in medical image analysis; it is a hard-tough problem in medical image processing and analysis. In this paper, we focus on fuzzy clustering algorithm. We analyze the principle of fuzzy clustering algorithm and its current development. The research work and innovations of this paper include:1. Separate medical image segmentation algorithms into 4 species after studying a large amount of literature. They are segmentation algorithms based on threshold, algorithms based on edge detection, algorithms based on regional characteristics, and algorithms combined with specific theory.2. Because the kernel-FCM (KFCM) algorithm is weak of robustness, we propose a new KFCM algorithm combined with Markov random field. KFCM algorithm doesn't consider neighborhood relationship among pixels, so it is very sensitive to noise. We put spatial constraint on KFCM with Markov spatial constraint field, so the algorithm gets better performance. The results indicate it could largely reduce misclassification rate of the white matter , gray matter and CSF than original KFCM algorithm.3. Propose a new automatic classification weighted fuzzy C-means (WFCM) algorithm for image segmentation. The classification number of traditional WFCM algorithm must be appointed in advance. We propose a new automatic algorithm. It automatically detects histogram peaks by smoothing image histogram in order to get classification number and initialize clustering center; then segment the images with WFCM. A large number of examples in this paper indicate that the new algorithm can automatically classify and effectively segment the images.