A New Approach: Generative Artificial Intelligence in Physiatry Resident Education

Abstract

Aim: This study assessed the effectiveness of ChatGPT-4o, an artificial intelligence (AI) platform, in creating a therapeutic exercises presentation for physiatry residents’ education. The aim was to compare the quality of content created by ChatGPT-4o with that of an expert, exploring the potential of AI in healthcare education. Material and Method: Both an expert and AI created 24 PowerPoint slides across six topics, using same reputable sources. Two other experts assessed these slides according to CLEAR criteria: completeness, lack of false information, appropriateness, and relevance and scored as excellent, 5; very good=4, good=3, satisfactory/fair=2, or poor, 1. Results: Interrater reliability was confirmed. Average scores (calculated from the two raters’ scores) for each topic were significantly lower for AI than for the expert, although whole presentation scores did not differ between the two. Overall scores (calculated from the average scores of all items) for each topic were good to excellent for AI, excellent for the expert. The overall score for whole presentation was good for AI, excellent for the expert. Highest ranks for individual criteria was relevance for AI, lack of false information for the expert. Some AI-generated elements were later integrated into the expert work, enhancing the content. Conclusion: ChatGPT-4o can generate effective educational content, though expert outperforms it, highlighting the need for professional oversight. Collaboration between humans and AI may further enhance educational outcomes.

Ähnliche Arbeiten