Medical Image Segmentation Based on GVF Snake Model

In this paper, the GVF snake model with the initial contour obtained by interactive method based on the adaptive threshold segmentation is proposed for the medical image segmentation. Firstly, input a line segment intersecting the desired object edge by user interaction. Find the max gradient magnitude where the cross-point is along the line segment. Then calculate the adaptive gray threshold and use growth algorithm to obtain the object area whose contour is the initial contour for the GVF model. Experiment shows that this algorithm can obtain the boundaries of the desired object from medical images quickly, accurately and reliably with little user interaction.