Generation of a Graph Representation from Three-Dimensional Skeletons of the Liver Vasculature

Extracting hepatic vasculature from three dimensional imagery is important for diagnosis of liver disease and planning of liver surgery. In this paper we propose a method for generation of 3D skeletal graph of liver vessels using thinning algorithm and graph theory. First of all, basic methodology in the proposed method is introduced. Secondly, the skeletonization method together with a pre-processing method on liver vessel images is employed to form liver skeleton image. Thirdly, a graphbased technique is then employed on the skeleton result to efficiently form the hepatic vessel system. The liver vessel tree generation method was evaluated on liver CT datasets to show its effectiveness and efficiency.