Mass classification in mammograms using neural network

Breast cancer is one of the main causes of death in women. An early detection is important to increase the survival rate. One of the common modality used for an early detection is mammogram. However, manual reading by the radiologists may affect the accuracy of the diagnosis. Hence, a Computer-Aided Diagnosis (CAD) system is developed as an aid to minimize the false alarm rate and to improve the diagnosis accuracy. The processes in a CAD system include pre-processing, segmentation, features extraction and classification. This paper investigates the classification of mass in mammograms using different sets of features with a back-propagation neural network as a classifier. The experimental results show that the performance of the classifier in terms of accuracy is higher with more hidden nodes in the neural network and more input features are fed to the classifier.