September 10, 2025Scientific ReportsOpen Access

Comparative evaluation of large language models performance in medical education using urinary system histology assessment

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Comparative assessment of LLMs in urinary histology shows varying performance, indicating tailored educational tasks are essential.

Performance of large language models varied significantly, with ChatGPT-o1 achieving the highest MCQ accuracy at 96.31%.
Clinical scenario generation was led by Claude-3.5, scoring 91.4% of the maximum, highlighting different strengths across tasks.
Evaluation employed two independent tasks—MCQ performance and scenario generation—using a multi-dimensional framework for quality metrics.
Findings underscore the necessity of aligning model capabilities with educational goals to enhance effective implementation in medical education.

Szabó et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark