Mammogram mass classification based on discrete wavelet transform textural features

This paper proposes an algorithm for early detection of breast cancer. This work incorporates Manual segmentation and textural analysis for the mammogram mass classification. Discrete Wavelet Transform (DWT) features act as a powerful input to the classifiers. A total of 148 mammogram images were taken from authentic mini MIAS database and under the supervision of classifiers, solid breast nodules were classified into benign and malignant. The classifiers used are K- Nearest Neighbor (K-NN), Support Vector Machine (SVM), Radial Basis Function Neural Network (RBFNN). It is found that RBFNN with DWT features outperform SVM and K-NN with 94.6% accuracy. The proposed system has a high potential for cancer detection from digital mammograms.