A Novel Method for Extraction of Spleen by Using Thin-plate Splines (TPS) Deformation and Edge Detection from Abdominal CT Images

The spleen is always hypertrophied in the patient with cirrhotic liver, which is regarded as one of complementary features for analysis of cirrhosis. Furthermore, segmentation of an abnormal spleen region based on CT or MR images is a crucial step in surgical planning. However, precisely carrying out this step remains a challenge due to either connectivities of the spleen to other organs or the shape, internal texture, and homogeneity of spleen that maybe extensively affected in case of liver diseases. In this paper, we propose a non-density based method for extracting the abnormal spleen region by edge detection processing. The accuracy of segmentation of spleen in 20 cases is the coincidence ratio 0.96 with the average error rate of 4.3%, and the result showed that our fully automatic method for the segmentation is effective and robust despite the presence of abnormal tissues within the spleen.