Skip to main content Skip to navigation

Designing VR Simulation System for Clinical Communication Training with LLMs-Based Embodied Conversational Agents

Project Overview

The document explores the innovative use of generative AI in education, particularly through the development of a Virtual AI Patient Simulator (VAPS) that combines Large Language Models (LLMs) and Embodied Conversational Agents (ECAs) to enhance clinical communication training in Health Professions education. It critiques the limitations of traditional simulation methods, underscoring the necessity for customizable and dynamic interactions that can authentically replicate real-world medical scenarios. By conducting user-centered research with Health Professions students, the design process identified specific challenges and insights that informed the creation of an effective virtual training tool. The findings suggest that VAPS not only improves engagement and learning outcomes but also prepares students more effectively for real-life clinical interactions, presenting a significant advancement in educational methodologies within the health sector. Overall, the integration of generative AI technologies in educational settings demonstrates the potential to transform training approaches, offering tailored experiences that enhance skill acquisition and professional readiness.

Key Applications

Virtual AI Patient Simulator (VAPS)

Context: Health Professions education focused on clinical communication training for students like nurses, physician assistants, and pharmacists.

Implementation: Developed through user-centered design, incorporating feedback from students and clinical faculty to create customizable VR scenarios powered by LLMs and ECAs.

Outcomes: Enhanced clinical communication skills, increased practice opportunities, and improved student engagement through realistic and dynamic interactions.

Challenges: Resistance to adopting new technologies, ensuring the realism of simulations, and the need for instructor training in using the system.

Implementation Barriers

Technological Barrier

The complexity of integrating advanced AI technologies like LLMs and ECAs into existing educational frameworks.

Proposed Solutions: Develop user-friendly design forms for educators and provide comprehensive training sessions on how to implement and use the technology effectively.

Cultural Barrier

Resistance from faculty and students to adopt VR as a training method due to unfamiliarity and comfort with traditional methods.

Proposed Solutions: Highlight the benefits of VR through pilot programs and gather testimonials from early adopters to encourage wider acceptance.

Project Team

Xiuqi Tommy Zhu

Researcher

Heidi Cheerman

Researcher

Minxin Cheng

Researcher

Sheri Kiami

Researcher

Leanne Chukoskie

Researcher

Eileen McGivney

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Xiuqi Tommy Zhu, Heidi Cheerman, Minxin Cheng, Sheri Kiami, Leanne Chukoskie, Eileen McGivney

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies