Segmentation of Colon Tissue in CT Colonography Using Adaptive Level Sets Method

In this paper, we introduce an adaptive level set method for segmenting a colon filled with air and opacified fluid in CT colonography. We first use simple thresholding method to remove most of the opacified liquid. Then, closed contours with manual seed initialization are propagated toward the region boundaries through the iterative evolution of an adaptive implicit function. During each iteration, information in each region is considered by estimating the parameters of probability density function (PDF). Finally, we evaluate accuracy of the proposed method by computing the overlaps between the manually segmented colon and the algorithm segmented results. The proposed method has been tested on 10 real CT colonography datasets, and the accuracy achieved is 96.06%.