Tetrolet transform based efficient brest cancer classification system

The leading cause of cancer mortality among women arises due to breast cancer. Digital mammogram plays an important role for cancer diagnosis. The early sign of breast cancer is the appearance of microcalcifications clusters on mammogram images. An efficient method to classify the microcalcification severity is presented in this paper. Tetrolet transform also named as adaptive Haar transform is utilized as feature extraction technique, in which energy features are extracted from the Tetrolet, decomposed mammogram. The extracted features are fed as input to the classifier. The classification of microcalcification clusters into benign or malignant is done by k nearest neighbor classifier (KNN). The proposed Tetrolet based classification of microcalcification approach achieves satisfactory performance than the conventional Haar transform.