Automatic segmentation of the sequential images of neural stem cell based on geometric active contour / 生物医学工程学杂志

ABSTRACT

A rapid and robust segmentation system is the base of the research on the change process of neural stem cells. In order to improve this system, against the segmentation of the multiple connected and blurred border cells, we present an improved geometric active contour algorithm based on the level set method and the curve evolution theory. This method can solve the topological change by itself and obtain the more real edges of the cells. This method has been adopted in the segmentation of the sequence of neural stem cell. The results prove the validity and veracity.