Multi-parametric MRI-based Radiomics to Predict Prognosis of Patients with Stages II-III Rectal Cancer

Motivation: Radiomics shows great potential in the prognostic assessment of rectal cancer (RC), but there are fewer studies on radiomics for prognostic prediction in stages II-III RC. Goal(s): To investigate the value of MRI-based radiomics in predicting the prognosis of stages II-III RC. Approach: Pre-treatment T2WI and ADC maps were employed to extract features and build predictive models. Survival curves were plotted by using the Kaplan-Meier survival analysis and compared with the log-rank test. Results: The radiomics model achieved the best efficacy in predicting disease free survival in patients with stage II-III RC and was also valuable in predicting overall survival. Impact: The radiomics models can be used to predict prognosis RC patients. It has significant advantages, individualization radiomics model allows for more precise risk stratification and tailored treatment planning, which can improve patient outcomes.