The images segmentation of liver malignant tumor based on CT images in HCC

Hepatocellular carcinoma (HCC) is one of the most common types of canceration. In this paper, several image segmentation methods are combined, improved and applied to the field of HCC image segmentation. The main techniques contain: 1. K-means clustering algorithm combined with region growing method. 2. Watershed algorithm based on foreground and boundary. 3. Region growing algorithm based on LBP and grey level. Via much research, it can be found out that the first two methods used in this paper have never been applied to HCC image segmentation. In addition, this paper also presents a new region growing method that based on LBP. In the part of the experiment, the applicability and difference of them will be discussed. What's more, this paper also discusses the improvement of these combination methods compared with the single methods. With comparing their segmentation result and accuracy, it can gets the best segmentation plan, which also lay the foundation for the next three-dimensional reconstruction of the tumor area.