Related Experiment Video
Updated: Jan 17, 2026

Estimate the Cognitive Load Using Electrocardiographic Measure: A Human-AI Collaborative Task
Published on: December 5, 2025
A feasibility study of an AI-based virtual human training design in suicide gatekeeper education
Pei-I Lee1,2, Chia-Yi Wu1,3
1School of Nursing, College of Medicine, National Taiwan University, Taipei, Taiwan.
Background:
Suicide gatekeeper training is a preventive approach that equips nonprofessionals with the knowledge and skills to recognize and minimize suicide risks. AI-based virtual human technology offers an innovative way to deliver the training via immersive and interactive learning experiences.
Objective:
The study evaluated the acceptability and practicality of using Virti® virtual human simulation technique to design a Question-Persuade-Referral communication training and explored university students' learning experiences and feedback of the design.
Design:
Both qualitative and quantitative findings reflected Virti® learning outcomes. User satisfaction and confidence of Virti and suicide gatekeeper knowledge were measured along with semistructured inquiries collecting the learning experiences. Virti® was implemented in a 3-hour gatekeeper training session, including classroom teaching (1.5h) and Virti® practice (1.5h) via mobile or tablet.
Results:
Fifty-three undergraduates participated in the study (average age 19.2). The post-training score of the item, "I become better understood that gatekeeper care is essential." was rated the highest (Mean=8.23, SD=1.68), while "confidence in acting as a suicide gatekeeper" rated the lowest (Mean=5.82, SD=2.41). Three themes emerged through student feedback: Enhancing communication skills and confidence; perceived value and educational potential of Virti®; and technical limitations and improvement of Virti®.
Conclusion:
The study validated that AI-based Virti® scenarios were feasible for suicide gatekeeper training and communication skills promotion. Addressing technical issues may further strengthen user confidence.

