Digital Alchemy: Shaping the Future of Psychiatry training with Artificial Intelligence
Location
Center for Advanced Learning and Simulation (CALS)
Publication Date
April 2025
Start Date
17-4-2025 8:00 AM
Description
As psychiatry adapts to rapid technological and cultural shifts, innovative tools are needed to train future practitioners. This study examines the use of AI-based patient models to simulate psychiatric encounters for medical students and residents. Two AI patients were developed using real clinical cases and refined with expert input, incorporating elements such as social determinants of health and family systems frameworks. Participants interviewed the AI patients in a 30-minute text-based session, then submitted diagnoses, treatment plans, and feedback on the realism of the encounter. Findings indicate that AI patients approximate real psychiatric presentations — with participant diagnoses clustering around the intended design. Residents and students differed in their perceived realism, with students generally rating the simulations more favorably. Qualitative analysis of interview transcripts revealed clinically relevant themes, including rapport-building strategies, depth of diagnostic questioning, handling of comorbidities, and integration of psychosocial factors. Residents tended to engage in more direct and educational interactions, aligning with their advanced training. Overall, AI patient models offer a promising supplement to psychiatric education, providing opportunities for realistic practice, immediate feedback, and exploration of complex clinical themes. Future directions include incorporating virtual reality and cultural psychiatry dimensions, while ensuring ethical use and continuous evaluation.
Recommended Citation
Roi, Cody DO; Moore, Shenitta; Maria, Shadia; and Varun, Amin, "Digital Alchemy: Shaping the Future of Psychiatry training with Artificial Intelligence" (2025). Dept. of Psychiatry Research Symposium. 11.
https://digitalscholar.lsuhsc.edu/psych_rd/2025/presentations/11
Digital Alchemy: Shaping the Future of Psychiatry training with Artificial Intelligence
Center for Advanced Learning and Simulation (CALS)
As psychiatry adapts to rapid technological and cultural shifts, innovative tools are needed to train future practitioners. This study examines the use of AI-based patient models to simulate psychiatric encounters for medical students and residents. Two AI patients were developed using real clinical cases and refined with expert input, incorporating elements such as social determinants of health and family systems frameworks. Participants interviewed the AI patients in a 30-minute text-based session, then submitted diagnoses, treatment plans, and feedback on the realism of the encounter. Findings indicate that AI patients approximate real psychiatric presentations — with participant diagnoses clustering around the intended design. Residents and students differed in their perceived realism, with students generally rating the simulations more favorably. Qualitative analysis of interview transcripts revealed clinically relevant themes, including rapport-building strategies, depth of diagnostic questioning, handling of comorbidities, and integration of psychosocial factors. Residents tended to engage in more direct and educational interactions, aligning with their advanced training. Overall, AI patient models offer a promising supplement to psychiatric education, providing opportunities for realistic practice, immediate feedback, and exploration of complex clinical themes. Future directions include incorporating virtual reality and cultural psychiatry dimensions, while ensuring ethical use and continuous evaluation.