Flexible ureteroscopy in renal anomalies: an explainable AI model for surgical outcome prediction from EAU endourology

Comprehensive Summary

This study investigates the use of flexible ureteroscopy (fURS) for treating kidney stones in patients with congenital renal anomalies, including horseshoe kidneys (HK), malrotated kidneys (MK), and pelvic ectopic kidneys (PEK). By applying machine learning (ML) and explainable AI (XAI) techniques, the authors developed predictive models to identify key factors influencing stone-free status (SFS) and other surgical outcomes. Findings showed that patients with MK achieved the highest rates of successful outcomes, while those with PEK experienced greater variability in surgical success. Across all groups, the presence of residual fragments was the most significant negative predictor of SFS, followed by longer operative times and advanced patient age.

Outcomes and Implications

The study shows that explainable AI may help doctors better prepare for surgery in patients with unusual kidney anatomy. By using AI to predict the chances of success and potential complications, surgeons may in the future plan procedures more effectively and make smarter, more consistent decisions, especially in cases where standard guidelines are limited or patient anatomy makes outcomes less predictable.