Research On Division Technology Of Brain Tumor Image

Medical image segmentation is the technique that divides the image into different regions which have different characteristics or extracts the interested objects from the image. The different regions or objects should be identity with the anatomical structure. Medical image segmentation is a significant and difficult problem in image processing and analyzing. Brain tumor image partition is always a research focus in medical image segmentation because of its important value of clinic application. We carry on a preliminary investigation into partition technology[0] of MRI brain tumor image after getting some professional knowledge about medical image segmentation algorithms and the special features of the brain tumor image.In this paper, we mainly studied the segmentation algorithms based on region and geometric active contour models. Geometric active contour model was implemented by the variation level set method. In the method based on region, we put forward a novel method to improve the PSO algorithm. The fitness functions were created according to the characteristics of brain tumor image and then the PSO algorithm has the ability of clustering as the FCM algorithm. The segmented results were satisfied by implementing the improved approach. In the method based on level set, C-V model, Li Chunming model and the improved C-V model have been studied. The improved C-V model takes advantage of both C-V model and Li Chunming model. It not only makes use of the region-gray and region-gradient information of the image but also is more controllability for regulating the parameters that inexistence in both models. In the end, by studying the estimative standards of the brain tumor image segmentation, we estimate and analyze the capability of MRI brain tumor image segmentation algorithms and models which involved in this paper.By implementing these segmented approaches involved in this paper, it is reasonable to draw a conclusion that these methods itself are uncertain. However, they could be improved to meet our demands.