Medical-AI trial: AI feedback vs. instructor feedback in dental education (preclinical simulation lab) (NCT07682116)

Dental skills are built before treating patients through preclinical training on "simulated teeth." Manual techniques such as filling and cutting improve through on-the-spot feedback from an instructor, but instructor time and numbers are limited, making thorough one-on-one guidance for every student hard to provide. AI could, in principle, return more frequent and uniform feedback on procedures during practice.

This trial compares AI-generated feedback with human-instructor feedback in a lab where third-year dental students perform restorations on simulated teeth. Per the registry summary, students trained with one of the feedback types, and effects were evaluated across several dimensions: (1) practical performance score, (2) state anxiety, and (3) student perception.

The primary endpoint is the practical performance score; the target is 140 participants, and the study is at the completed stage.

What stands out is that the study evaluates not only performance (the "result") but also psychological aspects such as anxiety and perception. In education, feedback quality affects not just learning but also how learners feel. AI-assisted instruction draws wide interest, yet whether it can supplement human teaching, and how learners receive it, needs evidence.

This positions the trial as one example of measuring effects when AI is introduced into medical and professional education (it does not establish the efficacy of AI educational support).