Cervical lymph node response to 131I therapy in differentiated thyroid cancer using radiomics and clinical features.

Background: To develop and validate a radiomics-based predictive model integrating clinical and imaging features to evaluate cervical lymph node response to I-131 therapy in differentiated thyroid cancer (DTC). Methods: A retrospective cohort of patients with pathologically confirmed DTC who underwent cervical lymph node dissection followed by 131I therapy was analyzed. Patients were stratified into Excellent and Non-Excellent response groups according to biochemical, imaging, and follow-up outcomes. Key inflammatory-related targets were initially identified through analysis of the GSE33630 dataset, and representative radiomics features potentially associated with inflammatory activity were subsequently extracted from pre-therapeutic CT images. Feature selection was performed using the least absolute shrinkage and selection operator (LASSO) and multivariable logistic regression to construct predictive models. Model discrimination, calibration, and clinical utility were assessed using AUC, calibration curves, Hosmer-Lemeshow test, and decision curve analysis. A nomogram was further developed to facilitate clinical interpretation. Results: The integrated clinical-radiomics model demonstrated favorable discrimination in both the training and validation cohorts, with good calibration and net clinical benefit across clinically relevant threshold probabilities. Conclusion: The proposed radiomics-assisted model shows promising ability to predict cervical lymph node response to 131I therapy in DTC, providing a potentially useful tool to support individualized postoperative management. Nevertheless, external validation and comparison with existing predictive strategies are required before routine clinical adoption.