An AI-Driven Virtual Patient Platform (CBT Trainer) for Training Cognitive Behavioral Therapy Practitioners Against...

Area of Science:

• Psychology
• Artificial Intelligence
• Medical Education

Background:

• Cognitive behavioral therapy (CBT) training faces challenges in supervised practice, feedback consistency, and competence assessment standardization.
• Traditional training methods often lack sufficient diverse case exposure and structured skill development opportunities.

Purpose of the Study:

• To evaluate the acceptability and feasibility of CBT Trainer (TTZ), an AI-powered virtual patient platform for CBT training.
• To assess the platform's ability to provide real-time feedback aligned with established competence frameworks for psychological practitioners.

Main Methods:

• A mixed-methods pilot study involving usability testing and a 2-stage approach with 59 participants from psychology training programs.
• Data collection included the System Usability Scale (SUS), platform engagement metrics, post-study questionnaires, comparative evaluation against role-play, and qualitative feedback.

Main Results:

• The CBT Trainer platform demonstrated excellent usability (mean SUS 82.20) and was engaged voluntarily by participants.
• Participants reported significant self-perceived competence improvement, especially in assessment and information gathering skills.
• Qualitative feedback highlighted the platform's strengths in competency-aligned feedback, accessibility, and providing a safe practice space.

Conclusions:

• The AI-based patient simulation (CBT Trainer) shows potential as a supplementary tool for CBT therapists, particularly for accessibility and immediate feedback.
• Further research with randomized controlled designs and objective competence assessments is recommended to validate these findings.