Wound Care XR

Designing an embodied LLM coach to support reflection within a VR simulation for chronic wound care education.

Role
XR Designer & Developer

Status
Ongoing — in development

Context
Koç University, KARMA XR Lab & School of Nursing, Koç University Hospital

Tools
Unity · Meta Quest 3 · MakeHuman (MPFB) · Blender · Oculus LipSync · ElevenLabs · Meta LLM Building Blocks, Open AI

Methods
Iterative XR prototyping, embodied LLM integration

Overview

Wound Care XR is a nursing education simulation developed at Koç University’s KARMA XR Lab in collaboration with the School of Nursing and Koç University Hospital. The original simulation, developed by Emir Bedirhan Aydın, introduces wound types through patient models, interactive interfaces, and audio instruction.

My contribution is the design and integration of Mira, an embodied LLM-based coaching agent. Mira connects learners’ spoken questions and wound-specific context to dialogue intended to encourage explanation and reflection on clinical reasoning.

Prototype demo — footage from an earlier development phase; the current implementation has since evolved.

My Contribution

I designed Mira’s character model, conversational behaviour, and interaction flow, and implemented the coaching system within the existing VR simulation. This work includes:

  • Coaching interaction: shaping the agent’s conversational role and responses to learners.
  • Context integration: connecting changing wound-specific information to the dialogue.
  • Embodied interface: creating the avatar and integrating speech interaction, lip-sync, and agent behaviour in Unity.

The wider team brings expertise in nursing education and clinical content; my work focuses on how the coach operates and interacts with learners within VR.

Research Direction

The coaching design centres on how dialogue can invite learners to explain their reasoning and reconsider their interpretations. Its contribution to reflective learning remains to be evaluated.

Building on this implementation, I am interested in how selected bodily, spatial, and interaction events could inform when and how Mira offers reflective guidance. This is a proposed research extension.

Research Status

The coaching system is under development and refinement. Study findings and publications will be added as the project progresses.

Related Work by the Team

Secure and accessible AI in nursing education: A modular agentic chatbot framework comparing ChatGPT-4o with an open-source LLM for chronic wound care

Gürler, N., Şengül, T., Bülbül, S. H., Yılmaz Akyaz, D., Güler, Ö., Yantaç, A. E., Aydın, E. B., & Kirkland-Kyhn, H.

Nurse Education in Practice, 93, 104789, 2026.

Note: This publication presents the AI framework developed by the broader research team.


Continue Exploring

Migration Experiences

Migration Experiences: Earlier work integrating an embodied LLM agent into an immersive social experiment — a related exploration of AI presence in VR research contexts.

Psychosis

Psychosis: Earlier work examining how immersive VR can support reflective learning in clinical education contexts.