A Hybrid Scheme for Mass Detection and Classification in Mammogram

Mammography is used to detect breast cancer in women. This paper provides methodology for breast mass detection and classification. It consists of four steps. First step is enhancement of mammogram image using homomorphic filtering. Second step is segment the enhanced image using local seed region growing algorithm. Third step is feature extraction using shearlet transform. Finally, the fourth step is classification of mass as either benign or malignant using support vector machine.