Enhancing medical education with AI: A comparative study of various large language models in radiology
PLOS One · 06.11.2026 · 21 (11) · e0360928
DOI: 10.1371/journal.pone.0360928 ↗Kurz gefasst
Vergleichsstudie in der Zeitschrift PLOS One, online erschienen am 06.11.2026. Verfasst von Noha Bakhsh und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Internal Medicine, Division of Radiology, Faculty of Medicine, King Abdulaziz University, Rabigh, Saudi Arabia. Hauptthemen laut MEDLINE: Radiology, Education, Medical, Artificial Intelligence. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: King Abdulaziz University.
Abstract (Originalsprache)
Objective
To compare how various Large Language Models (LLMs) perform when solving text- and image-based medical student-level radiology multiple-choice questions (MCQs).
Methods
We evaluated the accuracy of seven LLMs (i.e., ChatGPT-4o, ChatGPT-5, ChatGPT-5 Plus, DeepSeek-V3.1-Terminus, Microsoft Copilot, Claude 4, and Gemini 2.5 Flash) in solving 60 text-based and 60 image-based MCQs. We compared the seven LLMs’ accuracy across question types, domains, and topics.
Results
The mean accuracy across the seven evaluated LLMs was 79.9% (671/840 correct responses). Claude 4 achieved the highest observed accuracy at 86.7%, whereas Gemini 2.5 Flash and Microsoft Copilot recorded the lowest accuracy at 75.0% each. Accuracy was substantially lower for image-based questions than for text-based questions (64.5% vs. 95.2%). In the generalized estimating equations analysis, question type was significantly associated with response accuracy (Wald χ 2 (1) = 64.53, p < 0.001), as were question domain (Wald χ 2 (1) = 6.86, p = 0.009) and the question type × domain interaction (Wald χ 2 (3) = 80.26, p < 0.001). The omnibus effect of model was not statistically significant after accounting for these factors (Wald χ 2 (6) = 6.59, p = 0.361). Although Claude 4 achieved the highest observed accuracy, no pairwise model comparison remained statistically significant after Bonferroni correction.
Conclusions
Large Language Models are less accurate at solving image-based than text-based radiology questions, supporting their role as adjunctive educational tools that require faculty oversight rather than independent replacements for radiology instruction.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Internal Medicine, Division of Radiology, Faculty of Medicine, King Abdulaziz University, Rabigh, Saudi Arabia.
- Radiology Division, Department of Medicine, Faculty of Medicine, King Abdulaziz University, Rabigh, Saudi Arabia.
- Department of Radiology, King Abdulaziz University Hospital, Jeddah, Saudi Arabia.
ORCID-Profile: Noha Bakhsh, Ghofran A Ageely
Förderung
- King Abdulaziz University
Zitieren
APA
Bakhsh, N., & Ageely, G. A. (2026). Enhancing medical education with AI: A comparative study of various large language models in radiology. PLOS One, 21(11), e0360928. https://doi.org/10.1371/journal.pone.0360928
Vancouver
Bakhsh N, Ageely GA. Enhancing medical education with AI: A comparative study of various large language models in radiology. PLOS One. 2026;21(11):e0360928. doi:10.1371/journal.pone.0360928
BibTeX
@article{Bakhsh2026,
author = {Bakhsh, Noha and Ageely, Ghofran A},
title = {Enhancing medical education with AI: A comparative study of various large language models in radiology},
journal = {PLOS One},
year = {2026},
volume = {21},
number = {11},
pages = {e0360928},
doi = {10.1371/journal.pone.0360928}
}
EUVIMED