EARLY DETECTION AND CLASSIFICATION OF BREAST TUMOR FROM MAMMOGRAM IMAGES

A low dose X-ray technique of the breast known as mammography is popular due to its advantages over other imaging techniques. Even though only 2 percentage chance of being malignant radiologist usually recommend for a biopsy test. The unwanted biopsy test not only increase the anxiety among patient but also enhance the health care cost. The existing CAD system may misinterpret the suspicious lesion as false positive or false negative. To avoid such misinterpretation it is necessary to improve the existing CAD system such that it will accurately predict the suspicious lesion. This paper presents a novel approach which compares several hybrid image processing techniques to enhance the accuracy. Hybrid technique is defined as the technique which combine two or more techniques together. The accuracy of the proposed system is obtained as 95 percentages.