Liver Scar Tissue Image Based Cancer Diagnosis Method

This study proposes a segmentation method for scar tissue and normal cells of liver tissue image and designs a liver phase recognition system. In medical image of liver cancer, capturing characteristics of liver tissue to identify liver cancer phase is a new approach. Currently, doctors diagnose liver cancer by using their eyes, but it is inevitable that doctors may ignore some details because of fatigue. Therefore, we propose an automatic identification system for liver cancer phase to help doctors to determine liver cancer phase of liver cancer images. The liver cancer phase automatic Identification system proposed in this paper uses the image processing methods to segment the scar tissue and normal cells. This system is divided into two parts, one is the segmentation of scar tissue and normal cells, and another is automatically determining the liver cancer phase. Because of uneven distribution of light when shooting liver tissue image, there are light reflection surrounding the image and it will interfere segmentation. This study proposes a segmentation method which uses adaptive cross filter concept to solve this problem, and it can also reduce the computing time. This paper extracts scar tissue and normal cells from images of liver tissue and uses the different features in images of different phases as basis to judge the liver cancer phases. The experimental results show that, in segmentation of scar tissue and normal cells of liver cancer image, and the accuracy, recall rate, and F-value of liver cancer phase recognition system all achieve more than 92% recognition rate. Keywords—scar tissue, image segmentation, liver cancer, cancer diagnosis system.