A review of several trials looked at how artificial intelligence (AI) compares to traditional methods when teaching clinical skills. The study focused on learners practicing in simulation-based environments, such as those used for surgery or other invasive procedures.
Researchers found that AI-assisted training was more effective than traditional methods specifically for improving surgical and invasive procedural skills. However, the results showed no significant advantage for AI over traditional learning when it came to clinical assessment or diagnostic skills. Additionally, while vision-based deep-learning models showed clear benefits, large language models did not show a significant difference compared to standard training.
Learners who used AI reported higher satisfaction and lower stress levels than those in the control group. While procedure times and completion rates were similar across both groups, the findings suggest that AI is particularly useful for mastering complex physical tasks. Because this evidence comes from a meta-analysis of various trials, it provides a broad look at current technology but does not replace the need for expert guidance in medical education.
