Breast Cancer Detection Using Mammography

The presence of microcalcification clusters in mammograms contributes evidence for the detection of early stages of cancer. In this paper, a low-cost and high-speed neural network based breast cancer detection algorithm is presented. The microcalcifications are extracted with an adaptive neural network that is trained with cancer/malignant and normal/benign breast mammograms and a best accuracy rate of 99% for the classification of cancer/normal/benign is achieved. Key-Words: Breast Cancer Detection, Mammogram, Tumor Pattern Classification, Learning Vector Quantization, Medical Imaging