Prediction of Pelvic Lymph Node Involvement in Patients With Prostate Cancer Using an Artificial Neural Network

Abstract

We propose a training method using an artificial neural network (ANN) to model and predict the risk of pelvic lymph node involvement in patients with clinical T1c-T3 prostate cancer. The trained ANN calculates a prediction value for the risk of pelvic node involvement to aid clinicians in determining a risk threshold at which to justify the irradiation of pelvic lymph nodes. The modified Roach formula by Rahman et al used to predict the risk of pelvic lymph node involvement in patients with current-era prostate cancer will be used for training, cross correlation, and testing of the model.