A cad tool in the diagnosis of breast cancer using mammography

The advancements in medical imaging modalities provides ton of images for clinical diagnosis. Texture is one of the important information which can be extracted from the medical images for effective diagnosis and they are well analyzed by Computer Assisted Diagnosis (CAD) for the detection of abnormalities. In this work, a CAD approach to detect breast cancer as early as possible is presented. It has three stages. The initial identification of calcium deposit by thresholding and also the given mammogram is enhanced by a histogram based equalization approach in the preprocessing stage. The next stage is feature extraction where the Sub-band Spectral Histogram (SSH) is computed from the contourlet transform. In the final stage, the extracted SSH features in the previous stage is classified using SVM classifier. Results show that the presented CAD system on MIAS database has high diagnostic power with increased stability and acts as an effective second reader for radiologists for breast cancer diagnosis.