Evaluation of Medical Students' Experience with AI-Enhanced Clinical Skills Teaching and Curriculum Optimization Pathways

Abstract

Objective: To explore medical students' perceptions, attitudes, current status, and development prospects regarding artificial intelligence (AI)-enhanced clinical skills teaching, and to provide empirical evidence for optimizing the application of AI in clinical skills education. Methods: A questionnaire survey was conducted among students at a medical university. The questionnaire covered four modules: basic information, AI literacy cognition scale, current status of clinical skills teaching, and prospects of AI-empowered medical education. A five-point Likert scale was used, with statistical charts and analyses. Results: Sophomores and juniors exhibited higher AI literacy than freshmen, but all showed a pattern of "positive attitude yet insufficient mastery" (application attitude >4.0, technical mastery <3.7). Although AI teaching tools have achieved some penetration, they are mainly used for shallow applications such as intelligent Q&A, while deep applications like personalized learning recommendations remain underutilized and are difficult to integrate into practical courses. For optimizing basic clinical skills courses, students preferred innovative pedagogies such as PBL and CBL, and advocated for more hands‑on opportunities (e.g., community-based practice). Conclusion: Currently, AI in medical higher education faces a triple dilemma: "cognition-ability disconnect," "shallow application," and "theory-practice separation." Future AI-empowered clinical skills teaching should move toward personalized development, enhanced classroom flexibility, increased authenticity of clinical skills training, and integration with community service practice, thereby providing a theoretical basis for AI‑enhanced medical education.