Medical Image Segmentation Based On Fuzzy Clustering Algorithm Applied Research

Medical image segmentation is playing an increasingly important role in medical image analysis, and it has been widely used in all medical research fields. From medical research and clinical application point viewed, image segmentation is the basis of medical image processing, medical diagnosis and target identification. It applied in a lot of fields such as object extraction, quantitative analysis,3D reconstruction and so on. Medical image is uncertain and inaccurate, it is fuzzy in form. The complexity of human anatomy structure and various reasons caused by imaging process can all lead to the intrinsic uncertainty of the image. Fortunately, fuzzy set theory can resolve the problem effectively, so it has been widely used in medical image segmentation and become a hot topic in image segmentation. In this paper, we focus on fuzzy clustering algorithm. We analyze the principle of fuzzy clustering algorithm and its current development. The main research results can be concluded as follows:1. In this paper, the current development of medical image segmentation and the principles of main algorithms are deeply discussed. The main algorithms include segmentation algorithms based on threshold, algorithms based on edge detection, algorithms based on regional characteristics, and algorithms combined with specific theory. The algorithms are applied to segment MR brain image. The evaluation methods of segmentation are given and discussed.2. The theory of fuzzy clustering algorithm is studied, and many problems of the algorithm are researched, such as initialization of the number of clustering and cluster centers, setting of weight exponent and so on. The results of MR brain image segmentation based on FCM algorithm are given in this paper. We divide image into four parts, which include white matter, gray matter, cerebrospinal fluid and the background. The problems of FCM algorithm for image segmentation are analyzed.3. The conventional FCM algorithm is noise sensitive because of not taking into account the spatial information. To overcome the above problem, the research on FCM algorithm for image segmentation based on two-dimensional feature of gray and spatial information is done, and a FCM algorithm for image segmentation based on spatial constraints is discussed. Then we proposed a new fuzzy clustering algorithm for medical image segmentation based on spatial information. The algorithm is formulated by incorporating the spatial neighborhood into the objective function of standard FCM algorithm. A new idea is presented to update the cluster centers in pixel value and pixel neighboring value simultaneously, from which a new simple and effective two-dimensional objective function are derived to realize the segmentation. The new algorithm is applied to synthetic test image and MR brain image. It is shown from the experiments that the algorithm is more exact and'more robust than the conventional FCM algorithm.