Artificial intelligence (AI) is increasingly applied in endourology to enhance surgical planning, risk stratification, and outcome prediction. This systematic review evaluates AI-labelled prediction models predicting treatment-related outcomes in patients with urinary stone disease undergoing endourological procedures.
A systematic search of PubMed, EMBASE, Scopus, and Web of Science was conducted up to May 2025, following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Original studies assessing AI-labelled prediction models for clinical outcomes after ureteroscopy, retrograde intrarenal surgery, or percutaneous nephrolithotomy were included. Data were extracted on modelling approach, outcomes, prevalence, and reporting of discrimination, calibration, and clinical utility, as well as the validation strategy. The protocol was registered in PROSPERO (CRD420251090820).
Twenty-five studies were included, mostly retrospective and from Asian or European centres. Stone-free rate was the most frequent outcome. Infectious outcomes included febrile urinary tract infection, systemic inflammatory response syndrome, sepsis, fungal infection, and pyonephrosis, while other outcomes comprised operative time, radiation exposure, impacted stones, stone composition, transfusion, stenting, and broader perioperative outcomes. Across outcomes, discrimination was commonly reported, whereas calibration and formal assessment of clinical utility were less consistently reported; validation was predominantly internal, with limited external or temporal validation. One study reported a web-based implementation.
AI-labelled prediction models have been developed for multiple clinically relevant outcomes in urinary stone surgery. However, calibration, clinical utility, and robust validation remain inconsistently reported. Future work should prioritise transparent reporting of these elements, alongside multicentre external or temporal validation and more standardised outcome definitions, before routine clinical adoption.
European urology focus. 2026 Apr 28 [Epub ahead of print]
Natali Rodriguez Peñaranda, Frédéric Panthier, Stefano Di Bari, Nicolò Lugli, Marco Ticonosco, Antonio Botto, Francesco Di Bello, Maria Chiara Sighinolfi, Tommaso Calcagnile, Rosario Leonardi, Bhaskar Somani, Olivier Traxer, Salvatore Micali, EAU section of endourology
University of Modena and Reggio Emilia, Modena, Italy. Electronic address: ., Service d'Urologie, AP-HP, Hôpital Tenon, Sorbonne Université, Paris, France., University of Modena and Reggio Emilia, Modena, Italy., Department of Urology, AOU di Modena., Department of Neurosciences, Science of Reproduction and Odontostomatology, University of Naples Federico II, Naples, Italy., Department of Urology, Fondazione Policlinico Universitario Agostino Gemelli, IRCCS, Rome, Italy., ASST Lariana, Como, Italy., Kore University of Enna, Enna, Italy., Department of Urology, University Hospital Southampton NHS Trust, Southampton, UK.