Testicular salvage: using machine learning algorithm to develop a predictive model in testicular torsion

Study compares machine learning (Random Forest) versus classical Cox regression for predicting orchiectomy risk in testicular torsion patients. The ML model achieved superior performance (AUC 0.95, 92% sensitivity) using preoperative parameters including monocyte count, symptom duration, and prior Doppler studies, demonstrating potential for clinical decision support in acute scrotal emergencies.