The 2025 update on artificial intelligence models in pediatric urology: Results from the AI-PEDURO collaborative

The use of artificial intelligence (AI) applications is expanding rapidly. Here, we report an annual update of the AI-PEDURO (Artifical Intelligence in PEDiatric UROlogy) online repository (www.aipeduro.com), which reviews new AI models in pediatric urology, highlights emerging trends, and updates the living scoping review.

AI-PEDURO curates a living scoping review and online repository of models applied to pediatric urology. In this PRISMA-ScR concordant review, we searched MEDLINE, EMBASE, Scopus, and CINAHL from June 2024 to July 2025 for the 1-year update of AI-PEDURO. We synthesized eligible studies by study and model characteristics. Each study was appraised with the APPRAISE-AI tool, correlations between study quality and publication year were assessed, and all models were added to the continuously updated AI-PEDURO online repository.

Within the update period, 22 new studies were identified and added to the repository, bringing the total to 81. Hydronephrosis and pyeloplasty remained the most common topics, followed by vesicoureteral reflux and urinary tract infection. APPRAISE-AI assessment showed study quality was generally low to moderate. Based on the appraisal, model strengths were in clinical relevance and reporting quality, while weaknesses were methodological rigor, reproducibility, and error analysis.

Models continue to be applied to a wide range of clinical topics and are of generally moderate quality. While AI models have yet to make the transition into widespread clinical application in pediatric urology, there is significant promise from these pre-clinical models.