Write a PREreview

Background/Objectives: Vesicoureteral reflux (VUR) is an important pediatric urologic condition in which accurate grading on voiding cystourethrography (VCUG) may influence counseling, surveillance, antimicrobial prophylaxis, and surgical decision-making. Visual grading is partly subjective, particularly around the grade III-IV boundary. This review mapped the available evidence on AI-based pediatric VUR assessment on VCUG, with emphasis on diagnostic performance, clinician comparison, AI-assisted interpretation, and implementation gaps. Methods: A computer-assisted scoping evidence map was conducted using an exported screening dataset containing 500 records and an extraction dataset containing 13 included studies. Studies were classified as direct VCUG-based VUR grading, low/high-grade classification, AI-assisted clinician interpretation, or indirect clinical prediction. Because the exported files did not document independent duplicate screening, full-text verification, or database-specific deduplication, the synthesis was designed as an evidence map rather than a formal diagnostic meta-analysis. Results: Thirteen studies published between 2019 and 2025 were included. Reported AI performance was generally promising but heterogeneous. The strongest externally validated Deep-VCUG study reported external AUC values of 0.944 for unilateral reflux and 0.924 for bilateral reflux. A qVUR model achieved AUC 0.84 and improved grading reliability 3.6-fold. A multicenter VCUG-DAM study showed marked improvement in clinician AUC with AI support. Direct AI-versus-clinician evidence remained limited. Conclusions: AI may support pediatric VCUG interpretation by improving grading consistency and high-grade VUR recognition. However, current evidence does not establish autonomous superiority over radiologists. AI should currently be considered a decision-support tool rather than a replacement for expert clinician interpretation.