Breast cancer prediction using a neural network model

This paper reports results on using an artificial neural network (ANN) for predicting the estrogen receptor (ER) status, which is not always available, but has a place in therapy selection of breast cancer. Our results show that in more than two thirds of the cases, the ANN is able to predict the correct ER status. An optimum neural architecture was rescarched, and optimal outpoint for prediction selected on the basis of clinical data.