An SVM Approach to Liver Lesion Border Extraction for Liver Cancer Analysis

Authoritative analysis and prior identification of liver cancer is significant difficulty in the area of practical radiology. Doctors should know the feature of the tumor to provide effective treatment for the victim and helps doctors in further diagnosis. This paper intends the method for analyzing the liver under cancer positive environment. The proposed technique uses learning approach called Support Vector Machine (SVM) classifier to identify the liver from overlapped organs and tissues in CT image. It also uses systematic approach for liver lesion or tumor identification and extraction using image smoothing and refining method. The proposed technique constituted to extract the lesion with common reference point as backbone