Feature extraction and classification of breast cancer on dynamic magnetic resonance imaging using artificial neural network

A neural network system was designed to extract and analyze the quantitative data from time–intensity profile. These data was used to predict the outcome of biopsy in a group of patients with histopathologically proved breast lesions. The performance of an artificial neural network (ANN) was compared with radiologists using a database with 120 patients’ records each of which consisted of 14 quantitative parameters mostly derived directly from time–intensity profile. The network was trained and tested using the jackknife method and its performance was then compared with that of the radiologists in terms of sensitivity, specificity and accuracy using receiver operating characteristic curve (ROC) analysis. The network was able to classify correctly 107 of 120 original cases and yielded a better diagnostic accuracy (89%), compared with that of the radiologist (79%) by performing a constructive association between extracted quantitative data and corresponding pathological results (r=0.72, P<0.001).