DE03 Innovating dermatology education: a novel artificial intelligence-driven virtual patient simulation for medical students

Abstract

Artificial intelligence (AI) has been widely explored in dermatology for image analysis and diagnostic support; however, its application in medical education remains an emerging field. Exposure to dermatology in medical schools is often limited due to short placement durations and the opportunistic nature of clinical encounters. Simulation-based learning is well established in clinical training; however, its application within dermatology teaching remains limited. AI-enabled patient simulation offers a novel approach to addressing these educational challenges. The aim of this study was to design and implement AI-driven, high-fidelity simulation scenarios aimed at recognizing and managing acute dermatological emergencies. Simulation sessions were delivered to penultimate-year medical students during their dermatology placement. The simulation incorporated interactive voice technology, enabling real-time spoken dialogue between students and an AI-simulated patient. Custom prompts were developed to instruct the AI to act as a virtual patient. Students conducted history taking, performed physical examinations using a high-fidelity manikin, formulated differential diagnoses and proposed management plans. Student feedback was collected via postsession surveys. The AI tool was evaluated against predefined performance criteria including medical accuracy, response rate, consistency and realism. Scoring was completed independently by two examiners, with mean scores calculated for each parameter. Qualitative assessment of AI performance demonstrated the following scores: medical accuracy (83%), response rate (67%), consistency of patient history (83%) and realism of patient simulation (45%). Postsession evaluation revealed that 85% of students would value the inclusion of similar AI-led simulation teaching in future curricula. This project demonstrates an innovative application of AI in undergraduate medical education through the simulation of virtual dermatology patients. By shifting the role of AI towards experiential learning and patient interaction, this approach has the potential to enhance the accessibility and scalability of dermatology education.