Evidential reasoning-enabled deep learning for reliable treatment outcome prediction in cancer therapy

Treatment outcome prediction plays an important role in realizing personalized cancer therapy. In triple-negative breast cancer (TNBC), neoadjuvant chemotherapy (NAC) is widely used to downstage tumors and improve surgical outcomes. In head and neck cancer (HNC), early prediction of lesion progression can assist treatment planning. However, inter-patient heterogeneity in treatment response and tumor behavior limits the effectiveness of generalized treatment strategies. To address this issue, we developed an evidential reasoning rule-enabled deep neural network (ER 2 -DNN) for reliable outcome prediction in cancer therapy. The ER 2 -DNN combines Convolutional Neural Network (CNN) based image feature extraction with data augmentation, Monte Carlo dropout, test-time augmentation and evidential reasoning rule (ER 2 ) fusion for generating uncertainty-aware prediction. Across both TNBC and HNC datasets, the model showed consistent predictive performance with well-calibrated confidence estimates. The ER 2 -DNN provides a framework for supporting individualized oncology decisions through reliable image-based modeling. • We present a unified ER 2 -DNN framework for predicting treatment outcomes: pCR in TNBC and lesion progression in HNC, using only pre-treatment imaging data. • We incorporate the Evidential Reasoning Rule (ER 2 ) to perform structured, reliability-weighted prediction fusion, enabling better uncertainty estimation and model trustworthiness. • We systematically evaluate prediction calibration using Expected Calibration Error (ECE) and Maximum Calibration Error (MCE), also compare ER 2 to other alternative fusion strategies.