Predicting recurrent cases of intussusception in children after air enema reduction with machine learning models

This study developed machine learning models to predict intussusception recurrence in children after air enema reduction, with XGBoost achieving the best performance (AUC 0.718). Key predictive factors included air enema pressure, mass size, patient age, symptom duration, and absence of vomiting, offering clinicians a data-driven tool for risk stratification.