Review of Feature Extraction and Classification Methods used in CAD System

Abstract

Computer Aided Diagnosis (CAD) system provides significant benefit to the medical industry by automatically detecting various diseases like cancer, tumor, fracture etc. Different medical images like X-ray, Computed Tomography (CT), Magnetic Resonance Image (MRI) etc act as input for CAD systems depending upon the type of disease. Categorization into whether disease/abnormality exists or not in that medical image acts as an output. Feature extraction and classification are the two the most important steps in designing CAD system. Accuracy of the CAD system depends upon the method used for feature extraction along with classification from the medical images. In the present study, review of various feature extraction and classification methods is done which are proposed by researchers for the CAD system in the past few years. Primary focus of the study is to discuss the advantages and limitations of different CAD system which are proposed for detecting cancer and tumor.