Abstract: Cystoscopy is a commonly used diagnostic and therapeutic technique in urology. Training surgical residents to independently perform cystoscopy can effectively alleviate human resource shortages and improve the efficiency of medical services. We designed and implemented a cystoscopy training course employing a blended teaching approach of ″intelligent agent pre-tutoring + virtual simulation training + clinical case training″. Forty novice surgical resident physicians rotating in the Department of Urology at Shanghai Jiao Tong University School of Medicine, Ruijin Clinical Medical College from January to July 2025 enrolled in the study. After completing theoretical learning and interactive training guided by the agent, they proceeded to simulator operation training and assessment. Upon passing the simulator-based test, they received clinical practice training and evaluation. The teaching outcomes showed that all residents passed the simulator assessment with an operation time of (276.25±24.23) seconds, a simulator score of (94.65±3.04) points, and the number of injury and bleeding was 0(0) times. In real cases operation assessment, all residents were able to independently complete the examination under the guidance of a supervisor, with an operation time of (559.50±57.15) seconds and an examiner's score of (89.71±3.18) points. The teaching satisfaction survey indicated that the satisfaction rates for participants' interest in the teaching method, improvement in knowledge comprehension, and enhancement of operational confidence were 97.5%(39/40), 100.0%(40/40), and 95.0%(38/40), respectively. The design and implementation of this course can improve surgical residents' theoretical understanding and operational skills in cystoscopy, providing a new model for clinical skills training.

Key words: Artificial intelligence, Cystoscopy, Virtual simulation, Surgical Residents, Skill training