Deep Learning Procedures to Detect Breast Cancer 31 in Mammogram and MR Images and Relative Exploration of Images

Breast cancer is one of the deadly disorders found among female population widely world around. Cancer attacks human cells without any symptoms until it reaches advanced stage. So proper screening and diagnosing methods are needed to identify this disease in earlier stage. Several image processing techniques are in use but they have both advantages and limitations. This paper focuses two diagnosing methods currently in practice, namely Magnetic Resonance Imaging and Digital Mammogram method. For both the images, classification is done by deep learning methods like CNN model, Inception V3 model and Residual Network model and accuracy is shown highest in Mammogram images with 86.34% for Resnet50 model.