Background Microsurgery is associated with a steep learning curve that requires extensive training through supervised surgeries, cadaver practice, and simulations. The emergence of artificial intelligence (AI) in medical education offers a new potential avenue for microsurgery training by providing real-time feedback, performance analytics, and advanced simulation. This study aims to evaluate the scope, implementation, and outcomes of AI in microsurgical education for trainees across all levels. Methods A systematic review was performed in October 2024 following Preferred Reporting Items for Systematic Reviews and Meta-Analysis with extension for Scoping Reviews (PRISMA-ScR) guidelines. Four databases, including Embase, PubMed, Scopus, and Web of Science, returned 3,323 citations. Inclusion criteria were studies investigating the use of AI in the medical education of microsurgical trainees. Abstracts, commentaries, editorials, systematic reviews, and non-English studies were excluded. After two-stage screening, a total of 16 studies were included in this review. Results The assessed AI interventions appeared in the following number of studies: Computer Vision (n=13), Sensor-Driven Models (n=2), Classical/Statistical Machine Learning (n=4), Task-Specific Neural Networks (n=4), Transfer Learning of Neural Networks (n=3), Zero-Shot Inference of Pretrained Models (n=5), Augmented/Virtual Reality (n=5), and Anatomical Landmark Tracking (n=5). Upon full data extraction, three overarching themes were identified among studies 1) Objective Assessment of Microsurgical Skills, 2) Innovations in Microsurgical Education Materials, and 3) Improvement of Surgeon Workload and Performance. AI improved skill assessment (accuracy: 0.74–0.99), training, and workload optimization. AI-enhanced microsurgical training reduced training time (p=0.015), improved ergonomics, and minimized cognitive load, accelerating learning (β=0.86 vs. β=0.25). Conclusion Artificial intelligence has transformative potential in microsurgical education and practice, as emphasized by its capacity to enhance skill assessment, educational tools, and ergonomic support. Despite these enhancements, additional work is needed to address challenges such as data bias, standardization, and real-world implementation.
Raquepo et al. (Thu,) studied this question.