Research Of Level-set Segmentation Algorithms For Medical Image Based On ITK

This paper is about the research of level-set segmentation algorithm for medical image based on ITK. It has important significance and research value on three-dimensional segmentation and reconstruction of organ.In this research, at first, I studied a large number of filters provided by ITK, and designed an appropriate way to pretreatment the image and make it easy to segment, and unified human's subjective sensation and the treating processes. Then I studied and analyzed some segmentation algorithms based on level-set, especially, improved fast marching segmentation based on narrow-band, combined with segmentation model provided by ITK , which can achieve fast and accurate segmentation for medical image. Combination with Marching Cubes method, it can achieve three-dimensional segmentation and get good results.Level-set method can overcome limits to the algorithm when topology of image changes, and restore to the shape with the multiple structure effectively. Level-set method based on ITK may use many kinds of templates provided by ITK to realize the medical image segmentation effectively, which make segmentation image smoother, the boundary more accurate, the algorithm more efficiently.