Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 10, 2025BMC Medical EducationOpen Access

The performance of ChatGPT on medical image-based assessments and implications for medical education

View Full Paper
Ask AI
Bookmark
Share

Authors

YXYang XiangWCWei Chen

Discussion

Loading...

Member takes

Overview

Observational analysis evaluated AI performance on image-based USMLE items in medical education, indicating potential for instructional support.

Key Points

  • GPT-4o achieved an accuracy of 89.5% on medical image questions, while GPT-4 scored 73.4%.
  • Prompt engineering was applied to enhance responses from both GPT-4 and GPT-4o in image assessments.
  • The study explored the application of these generative AI models in case-based teaching scenarios in medical education.
  • Human oversight is essential to ensure accuracy in AI-generated content, despite the promising performance of the models.

Cite This Study

Xiang et al. (2025) studied this question.

synapsesocial.com/papers/68c24317b210217d647a541ehttps://doi.org/10.1186/s12909-025-07752-0
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1A multi-model longitudinal assessment of ChatGPT performance on medical residency examinations2025 · 5 citations
  2. 2Evaluating the Potential and Accuracy of ChatGPT-3.5 and 4.0 in Medical Licensing and In-Training Examinations: Systematic Review and Meta-Analysis2025 · 20 citations
  3. 3Performance of GPT-4o and o1-Pro on United Kingdom Medical Licensing Assessment-style items: a comparative study2025
  4. 4Evaluating the Potential and Accuracy of ChatGPT-3.5 and 4.0 in Medical Licensing and In-Training Examinations: Systematic Review and Meta-Analysis (Preprint)2024
  5. 5Comparative performance of ChatGPT, Gemini, and final-year emergency medicine clerkship students in answering multiple-choice questions: implications for the use of AI in medical education2025 · 11 citations