A Method to Develop Reliable “Ready to Use” Graphical Tools Based on Artificial Neural Networks for the Prediction of Toxicities After High Dose Prostate Radiation Therapy

To study a reliable application of artificial neural network (ANN) classification for the prediction of any selected acute or late toxicity endpoint following prostate cancer radiation therapy (RT) and to develop a method to achieve “ready to use” paper tools based on the obtained ANN model. The method was tested on population of 550 men recruited in a large multicenter prospective trial. Clinical information was available on comorbidities,...