Breast Cancer Classification Using Image Processing Technique and Support Vector Machine

Abstract

In women world, according to Cancer Prevention and Control (CPC) breast cancer is the second largest cause of death next of lung cancer but if it is diagnosed early it is also one of the curable cancer. Radiologist reads mammography image manually which is a tedious and confusing task making them over sight errors to fail to detect the cancer. This research aims to investigate the possibility of detecting and classifying breast cancer using image processing technique. For better image detection, first the quality of the input mammography image improved at the preprocessing stage by removing noises and enhancing the contrast of the image. Next, thresholding, which is one of the level set methods, is used to obtain the region of interests (ROIs) for each mammogram. Refining the segmentation process is achieved using image masking and image filtering technique. Then, geometrical features are extracted from the ROIs. Finally, Support Vector Machine (SVM) is used as a classifier to distinguish mammograms as normal and abnormal. MATLAB environment is used to train and test the proposed approach using percentage -split (70% for training and 30% for testing) evaluation method. In this study, the experimental result shows that 82.14% and 78.95% overall accuracy achieved using level set method with linear SVM classifier which are applied on 98 (lcc and lmlo) images and also on 193 (all images) local dataset mammography images respectively.