828 Real-Time Artificial Intelligence Instruction in Comparison to Human Expert Instruction in Surgical Technical Skills Teaching – a Randomized Controlled Trial

Abstract Aim Surgical simulators equipped with artificial intelligence (AI) systems provide risk-free training on realistically simulated complex patient cases, objective performance assessment, and tailored error avoidance feedback. This study compared the efficacy of real-time AI instruction human instructor-mediated in surgical technical skills teaching. Method Medical students (n = 98) were randomly allocated into three feedback groups. Everyone repeated a virtual brain tumour resection five times. The first repetition was done without receiving to assess their baseline performance. Students Group 1 did not receive feedback, they shown visual only after each procedure based expert benchmarks. 2 instructed by system. After task, error-video clips generated this system alongside expert-level demonstrations relating error. 3 instructors during tasks. summarized areas improvement demonstrated correction techniques. All participant data scored using composite-score. Statistical analysis conducted compare learning from last task repetition. Results significantly improved score third second repetition, respectively (p<0.01, p 0.01) achieved higher scores than those fifth (p<0.01). Conclusions Artificial may competency-based curricula efficient teaching methodologies.