Authors: (including presenting author): :
Li PC(1), Law KL(2), Wong MC(3), Cheung SK(1), Kan GL(1), Chu KL(4)
Affiliation: :
(1)The Hong Kong University of Science and Technology (Guangzhou), (2)Kowloon East Cluster Pain Management Centre, United Christian Hospital, (3)Nursing Services Division, United Christian Hospital, (4)Department of Anaesthesiology, Pain Medicine and Operating Services, United Christian Hospital
Keyword 1: :
Simulation-based Training
Keyword 2: :
Epidural Analgesia Care
Keyword 3: :
AI-enabled Healthcare Training
Introduction: :
Epidural analgesia care is a crucial clinical procedure. It requires professional knowledge and proficient psychomotor skills, particularly catheter removal and entry-site dressing. However, current training approaches often lack simulation of challenging situations such as high resistance during catheter removal.
Objectives: :
This project aims to develop an intelligent simulation training platform that combines realistic, challenging situations with interactive software to improve the training effect and quality of care for patients receiving epidural analgesia.
Methodology: :
We collaborated with anesthesiologists, pain management nurses, and multidisciplinary researchers to develop a system with a portable hardware simulator and a tablet-based software platform. With this iterative co-design approach, we developed a hardware mimicking the human lower back, an LED array provides entry-site bleeding cues, and a motor-driven winding mechanism and servo structure generate variable catheter-withdrawal resistance. The software integrates camera-based hand and gesture recognition with a large language model (LLM). Key features include: (1) a simulation training module for epidural analgesia care (catheter removal and entry site dressing change), using gesture recognition for performance assessment and event-triggered guidance; (2) an e-learning module with multimedia materials and embedded quizzes; and (3) an LLM-powered AI Nursing Mentor to provide real-time instructional support and answer trainees’ questions.
Result & Outcome: :
Following a co-design session on 14 July 2025, a prototype was evaluated on 8 January 2026. Seven anesthesiologists, operating room and pain management nurses reported a positive first impression (3 “very positive,” 4 “somewhat”) and showed rapid functional understanding (4 “immediately,” 3 “within moments”) of the platform. Users rated intuitiveness (4.00 ± 0.89) and visual clarity (4.20 ± 0.45) on a 5-point scale, with no technical crashes reported. Key strengths included realistic haptic resistance and system portability. Primary improvement requests focused on adding diverse resistance scenarios, improving gesture recognition function for dressing changes, and implementing instructor controls for multi-trainee monitoring. This work demonstrates the feasibility of a simulation platform for comprehensive epidural analgesia care. Early feedback confirms its usability and stability, and identifies priorities for future iterations, including expanding scenario diversity, enhancing gesture recognition for complex workflows, and advancing instructor-facing tools.
