RESIDENT: Real-Time Al Voice Simulation Platform for Psychiatric OSCE Training

Evaluate the session

Presenting Author(s): Abhinav Pillai

Co-Author(s): Rufina M.J. Kim, MSc; Tan Li; Benjamin Rosen, MD; Thomas Stark, MD; Martin Vetter, MD; Yanbo Zhang, MD, PhD

Date and time: 20 Mar 2026 from 15:20 to 15:35

Location: Saddleback & Glacier  

Learning Objectives

1. Describe the design and functionality of the real-time Al voice simulation platform, including its use of OpenAl's voice API, psychiatric case staging, emotional responses, and automated examiner feedback;
2. Explain how real-time Al voice technology can be applied to simulate psychiatric OSCE encounters, including its ability to generate dynamic patient interactions and examiner feedback;
3. Identify the potential educational benefits and scalability of Al-generated OSCE simulations, including standardized encounters, on-demand practice, and support for building clinical interviewing and diagnostic skills; and
4. Explore limitations of the current technology and future directions.

Abstract

Background: Objective Structured Clinical Examinations (OSCEs) are essential for assessing clinical competencies in psychiatric training, yet provide limited opportunities for repeated practice due to resource and scheduling constraints. 

Methods: We developed an innovative training platform utilizing OpenAl's real-time voice API to simulate encounters for OSCE preparation. The system enables residents and medical students to engage in real- time, voice-based clinical interviews with Al-generated examiners presenting various psychiatric conditions and situations. The platform incorporates authentic OSCE staging, emotional responses, and challenging clinical scenarios commonly encountered in psychiatric OSCEs, along with an Al examiner component that analyzes interviews and provides structured feedback. 

Results: Our platform provides on-demand, standardized practice opportunities for learners to practice psychiatric interview skills, receive immediate and personalized feedback, and refine their clinical skills without the barrier of scheduling constraints. The voice-based interaction attempts to mimic the real-life conversations that happen during actual OSCE encounters, capturing details like pauses and reactions.

Conclusion: This Al-powered training tool has the potential to address critical gaps in psychiatric OSCE preparation by offering on-demand simulations that can be tailored to learners' needs. This platform can easily be scaled given its virtual nature. The tool demonstrates potential for enhancing clinical interview skills, building diagnostic confidence, and preparing trainees for high-stakes assessments while reducing dependence on human standardized patients and faculty time. 

Literature References

1. Yamamoto, A., Koda, M., Ogawa, H., Miyoshi, T., Maeda, Y., Otsuka, F., & Ino, H. (2024). Enhancing Medical Interview Skills Through Al-Simulated Patient Interactions: Nonrandomized Controlled Trial. JMIR Medical Education, 10, e58753. doi:10.2196/58753 
2. García-Torres, D., Vicente Ripoll, M. A., Fernández Peris, C., & Mira Solves, J. J. (2024). Enhancing Clinical Reasoning with Virtual Patients: A Hybrid Systematic Review Combining Human Reviewers and ChatGPT. Healthcare, 12(22), 2241. doi:10.3390/healthcare12222241. 
3. Jo, Y.-W., Lee, M., & Yang, H.-J. (2025). Large Language Model-Based Virtual Patient Simulations in Medical and Nursing Education: A Review. Applied Sciences, 15(22), 11917. doi:10.3390/app152211917. 
4. Kelly, S., Smyth, E., Murphy, P., & Pawlikowska, T. (2022). A scoping review: Virtual patients for communication skills in medical undergraduates. BMC Medical Education, 22, Article 429. doi:10.1186/s12909-022-03474-9. 
5. Chaby, L., Benamara, A., Pino, M., Prigent, E., Ravenet, B., Martin, J.-C., Vanderstichel, H., Becerril-Ortega, R., Rigaud, A.-S., & Chetouani, M. (2022). Embodied virtual patients as a simulation-based framework for training clinician-patient communication skills: An overview of their use in psychiatric and geriatric care. Frontiers in Virtual Reality, 3, 827312. doi:10.3389/frvir.2022.827312.