Active Contour Model Applied In Tumor Image Segmentation

Now tumor becomes fatal illness threating healthy of human being, millions of people in the world died of tumors every year. Extraction the contour of tumor and adjacent organs is very important for diagnosis and treatment of tumor. Aiming at the issue, this paper proposes approach for tumor image segmentation based on active contour, which can provide more effect and accuracy results for diagnosis and treatment of tumor.In this paper, the developing active contour model theory at home and overseas is studied and analyzed. The application in the field of tumor image segmentation based on active contour model theory is proposed as the main study objective. This paper analyzes the theory and describes the mathematic model and numerical solution method in detail; it also discuss the application in tumor segmentation and organ segmentation.The main works in this paper are as follows:1. Achievements and actuality of medical image segmentation is reviewed and summarized in this paper, and methods for medical image segmentation are classified. The basic theory of active contour model is studied in this paper. It includes the representation of the model, the composition of the model's energy function and it's minimum, and the limitations in application. It presents the nature of the energy components in the model and analyzes the model's activity affected by the energy components.2. Basic theory of parametric active contour model and its numerical solution is discussed. Classical external force field is replaced by GVF model for capture range enlarging and sensitivity of contour initialization, which can result in better boundary concavity convergence of medical image. Basic theory of FCM algorithm is discussed and FCM algorithm is used for previous segmentation of medical image with weak edges, artifacts and local gradient maximum regions. The result of FCM can drive active contour and control the speed and direction of it, so it can make curve shrink to the boundary of tumor or important organ.3. Basic theory of geometric active contour model is studied. With the virtue that can change the topology freely, geometric active contour model can be used for medical image processing based on curve evolvement and level set method. Narrow band method can be used for image segmentation with geometric active contour model that has large computational magnitude. Theory and solution approach of watershed algorithm is...