An artificial neural network model was developed to predict key radiobiological parameters in head and neck squamous cell carcinoma patients undergoing radiation therapy, using factors like tumor specifics and radiation doses.
The study utilized two datasets containing information on 31 and 215 patients, extracting relevant data to train a single-layer neural network that predicts tumor control probability and complications.
The model achieved high prediction accuracy, with values around 0.89 during training, indicating its potential as a useful tool for clinical treatment planning in cancer management.