AI-AUGMENTED VIRTUAL OPHTHALMOLOGY MICRO-CLERKSHIP TO IMPROVE DIAGNOSTIC COMPETENCE IN MEDICAL STUDENTS: A CONTROLLED COHORT STUDY

Background: Undergraduate ophthalmology exposure is frequently time-constrained, while image-centric competencies (e.g., slit-lamp pattern recognition, OCT interpretation) are increasingly central to modern practice. Digital approaches—virtual clinics, flipped learning, and simulation—have shown educational benefit, but pragmatic, assessment-anchored implementations remain underreported in many curricula. Methods: In a controlled cohort study, 175 medical students were allocated to an AI-augmented virtual micro-clerkship (Intervention, n=88) or traditional tutor-led seminars (Control, n=87) for 2 weeks. Primary outcome was post-module ophthalmology OSCE score (0–100). Secondary outcomes included OCT interpretation accuracy, diagnostic reasoning rubric score, time-to-diagnosis, and learner-reported measures. Results: The intervention group achieved higher OSCE performance (84.6±11.3 vs 77.2±12.8; p<0.001) and higher OCT accuracy (78.4%±12.6 vs 69.1%±13.9; p<0.001), with moderate effect sizes. Conclusion: A short, AI-supported virtual micro-clerkship can measurably improve ophthalmology performance while maintaining feasibility for crowded curricula.

Перевод пока недоступен