Breast cancer diagnosis based on a suitable combination of deformable models and artificial neural networks techniques

Abstract

According to the World Health Organization (WHO) breast cancer is the most common cancer suffered by women in the world, which during the last two decades has increased the women mortality in developing countries. Mammography is the best method used for screening; it is a test producing no inconvenience and with small diagnostic doubts of breast cancer since the preclinical phase. For this reason, unfailing Computer-Aided Diagnosis systems for automated detection/classification of abnormalities are very useful and helpful to medical personnel. In this work is proposed a novel method that combines deformable models and Artificial Neural Networks among others techniques to diagnose diverse mammography abnormalities (calcifications, well-defined / circumscribed masses, spiculated masses, ill-defined masses, architectural distortions and asymmetries) as benign or malignant. The proposed algorithm was validated on the Mammographic Image Analysis Society (MiniMIAS) database in a dataset formed by 100 mammography images, which were selected randomly. © Springer-Verlag Berlin Heidelberg 2007.