Development and Evaluation of an Artificial Intelligence-Based Virtual Patient Chatbot for Diagnostic Education in Korean Medicine

Objectives: This study aimed to develop an AI-based virtual patient chatbot to address the limitations of clinical practice and CPX(Clinical Performance Examination) in Korean medicine diagnostics education and to evaluate changes in students’ learning motivation and usability after its application in class.Methods: The virtual patient chatbot was designed and developed based on the ADDIE model. It consisted of stages including history taking, test selection, differential diagnosis, final diagnosis, and feedback, and was implemented in a web-based environment. Twenty-four students from a college of Korean medicine participated in pre- and post-intervention surveys. Learning motivation in diagnostics was measured using items based on the Motivated Strategies for Learning Questionnaire (MSLQ), and pre-post differences were analyzed using the Wilcoxon signed rank test. Usability was evaluated in terms of ease of use, usefulness, and satisfaction.Results: The learning motivation score in diagnostics significantly increased after virtual patient chatbot-based learning compared to baseline (p=0.003). Usability evaluation indicated overall positive perceptions (3.81±0.72), with usefulness (4.11±0.66) rated the highest among the subdomains. In addition to quantitative findings, qualitative responses revealed that students recognized improvements in structuring question flow, understanding the diagnostic process, and engaging in repeated self-directed practice.Conclusions: The AI-based virtual patient chatbot demonstrated measurable improvements in learning motivation and was positively perceived as a structured learning tool in diagnostics education. By integrating history taking, diagnostic test selection, differential diagnosis, and feedback within a single platform, the chatbot offers a standardized and repeatable model to supplement CPX training. These findings support the potential of AI-driven virtual patients as scalable educational tools in competency-based Korean medicine education.