Analysis of Liver Cancer Using Data Mining SVM Algorithm in MATLAB

Abstract

Liver cancer is one among the normal types of cancer. Detection and determination of liver tumor at early stage are vital. The vast majority of the cancer passings can be anticipated by early detection, determination, and compelling treatment. It is required to fragment the liver tumor from the medical images for tumor analysis. A robotized framework is proposed for segmentation and classification of liver tumor which is an effective and simple to utilize technique. The proposed framework comprises of PreAprocessing, segmentation, postAprocessing, and a last classification as benign and malignant. Amid the preAprocessing stage, the image is resized to 256 × 256. In the segmentation stage, level set strategy is connected for sectioning the suspicious area. In postAprocessing stage, the district of intrigue is acquired from the first image. At long last the Pseudo Zenerike minute and GLDM is utilized for highlight extraction from CT image. These components are given as contribution to the SVM for classification of tumor as benign or malignant. The SVM is prepared utilizing four images. The proposed framework can accomplish precision rate of 86.7%.