Medical Image Segmentation Based On Contour Model

Segmentation of medical images plays a key role in medical imaging technology, and the effect of which will impact the results of computer assisted diagnosis. Recent years, a variety of algorithms has been proposed, most of which,however, concentrate on a specific issue.This thesis concentrates on the research on segmentation of medical imagesbased on active contour models. The concept of active contour models was creatively proposed by Kass M、Witkin A、Terzopoulos D in1987. After entitled "Snakes: Active Contour Models" paper published, Snakes segmentation model soon became the most successful of the most active research areas. Currently active contour segmentation model used to describe the dynamics of the target profile has been successfully used in object recognition, computer visionand biomedical image area.In this paper, we summarize and review the existing medical image segmentation methods, analysis of the corresponding advantages and disadvantages. Start from the dynamic contour model, introduced the classic Snakes model, and discussedthe model of two classic Snakes improved model (GVF Snakes and Ballon Snakes), summed up the advantages and disadvantages of these two models. for the traditional Snakes model, this paper discusses the vector field convolution model, which in the capture range, the recessed area of the convergence zone, the initial contour, noise immunity and the complexity of the algorithm has done a comprehensive improvement. Snakes model initialization for a long time, the paper discusses the inverse Poisson distribution of gradient methods, the introduction of automatic initialization method, as an independent process, through the precise contours of the automatic initialization and shorten the iteration time, and then Snakes model to reduce theconvergence time.The experiments on medical images to further verify the feasibility of this method. Finally, the paper points out further research.