Trust-VR: Feasibility of AI Architectures for Virtual Pediatric Patients
Context
Managing distressed pediatric patients in clinical environments is a challenging yet critical skill for healthcare professionals. Patients exhibit diverse emotional responses—ranging from anxiety and shyness to outright resistance—making it essential for clinicians to adapt their approach. Traditional training methods often lack the realism and variability needed to prepare professionals for these high-stakes interactions, particularly when it comes to emotional and behavioral dynamics in children.
Motivation
A VR solution featuring animated child patient avatars with dynamic, AI-driven behavior can provide a realistic, hands-on training experience. By engaging with these avatars in a safe, repeatable, and immersive environment, healthcare professionals can practice communication, de-escalation, and trust-building techniques. This project focuses on a feasibility study comparing AI control architectures, such as behavior trees and reinforcement learning–based controllers, to determine which is most suitable for modeling realistic, interactive child behaviors in pediatric VR training.
Goal
This project shall conduct a feasibility study identifying which AI architecture is best suited for animating virtual children in VR-based pediatric training (e.g., behavior trees, reinforcement learning, or hybrid approaches). The project will prototype and evaluate different control schemes for child avatars that exhibit diverse emotional and behavioral responses, such as fear, pain, withdrawal, and cooperation. The outcome will be design and implementation guidelines for the Trust-VR platform, indicating whether behavior trees, reinforcement learning tools, or a combination should be used to support realistic, adaptable training scenarios.
Requirements
The student should have:
- Familiarity with Unreal Engine (C++ or Blueprint) and its AI tools (e.g., behavior trees, blackboards, perception systems).
- Basic understanding of AI decision-making frameworks (behavior trees, finite-state machines, and introductory reinforcement learning).
- Interest in healthcare or medical simulation contexts, especially pediatric care and communication training.