A Bayesian optimization tunning integrated multi-stacking classifier framework for the prediction of radiodermatitis from 4D-CT of patients underwent breast cancer radiotherapy

A novel multi-region dose-gradient-based Bayesian optimization tunning integrated multi-stacking classifier framework can achieve a high-accuracy prediction of symptomatic RD 2+ in breast cancer patients than any other single deep machine learning algorithm.