Prediction error comparison between the previous model, the proposed model and proposed model with data augmentation. The proposed models are significantly better at predicting all OARs, including brainstem (P < 0.001), cord (P < 0.001), larynx (P = 0.004), mandible (P < 0.001), pharynx (P = 0.001), oral cavity (P = 0.011), and parotid (P < 0.001). [Color figure can be viewed at wileyonlinelibrary.com]