Solving Mammography Problems of Breast Cancer Detection Using Artificial Neural Networks and Image Processing Techniques

Abstract

In this paper, we propose a complementary technique of breast c a nc er diagnosis that covers five stages of breast cancer detection based on mammography, which solves many of the problems found otherwise. We also show a very large area w h e r e many methods and techniques can be successfully merged in order to obtain a useful result for human use. These include scaling of the image, removing small objects, smoothing, extracting features, ROI extraction and many image processing techniques. Besides, neural networks are used here to train the system to detect cancer according to the dataset. This combination of multiple techniques can solve problems of the breast cancer detection with a high degree of accuracy. Examples and comparisons are given to illustrate and prove this method.