Artificial Intelligence in Surgical Feedback: A Systematic Scoping Review

Artificial intelligence (AI) is poised to transform surgical education, particularly in providing objective, scalable feedback. This study aims to systematically evaluate the current landscape of AI applications in delivering procedural feedback to surgical learners.

A systematic scoping review of literature published between 2015 and 2025 was conducted using search terms related to AI, surgical education, and feedback. Following PRISMA guidelines, studies were included if they detailed the implementation of an AI-based system for surgical or procedural feedback. Retrieved articles underwent title, abstract, and full-text screening to determine final eligibility.

The initial search yielded 2374 unique studies, with 18 meeting the final inclusion criteria. A risk of bias analysis revealed that twelve of these studies (67%) presented a moderate risk of bias. The included studies spanned multiple specialties, including general surgery (33%), neurosurgery (17%), and otolaryngology (6%). The efficacy of AI feedback was primarily assessed through performance metrics against expert benchmarks and subjective learner surveys. Analysis revealed that AI-driven feedback was, in several studies, found to be equivalent or superior to expert human instruction in simulation settings. While most studies (n = 14) utilized post-task feedback, real-time AI models demonstrated the potential for faster error correction. The majority of studies focused on trainees (n = 10), though AI was shown to be a valuable educational tool for learners at all skill levels. Five studies were randomized controlled trials, indicating a trend toward more rigorous study designs.

AI represents a promising and significant supplement to traditional procedural feedback methods. Future development, particularly for applications in otolaryngology, will depend on interdisciplinary collaboration to overcome barriers such as cost and user trust.