AI performance in emergency medicine fellowship examination: comparative analysis of Chatgpt-4o, Gemini 2.0, Claude 3.5, and Deepseek R1 models
Large language models can achieve high accuracy rates for knowledge and clinical reasoning questions in emergency medicine but show differences in terms of response consistency and hallucination tendency. While these models have significant potential for use in medical education and as clinical decision support systems (CDSS), they need further development to provide reliable, up-to-date, and accurate information.
Authors
- Bensu Bulut (ORCID: https://orcid.org/0000-0002-5629-3143)
- Mustafa Sait Gönen (ORCID: https://orcid.org/0000-0002-1089-623X)
- Mehmet Yortanlı (ORCID: https://orcid.org/0000-0002-6744-2423)
- Hüseyin Mutlu (ORCID: https://orcid.org/0000-0002-1930-3293)
- Ayşenur Gür (ORCID: https://orcid.org/0000-0002-9521-1120)
- Murat Genç (ORCID: https://orcid.org/0000-0003-3407-1942)
- Medine Akkan Öz (ORCID: https://orcid.org/0000-0002-6320-9667)
- İshak Şan (ORCID: https://orcid.org/0000-0002-9658-9010)
Institutions
- Aksaray University (TR)
- Ministry of Health (TR)
- Bilkent University (TR)
- University of Health Science (KH)
- State Hospital (GB)
- Konya Numune Hastanesi (TR)
- Gülhane Askerî Tıp Akademisi (TR)
- Memorial Ankara Hospital (TR)
- Ankara Numune Eğitim ve Araştırma Hastanesi (TR)
- Başkent University Hospital (TR)
- Sağlık Bilimleri Üniversitesi (TR)
- İstanbul Kanuni Sultan Süleyman Eğitim ve Araştırma Hastanesi (TR)
- Etimesgut Asker Hastanesi (TR)
- University of Health Sciences Antigua (AG)
Publication Details
- Journal
- TURKISH JOURNAL OF MEDICAL SCIENCES
- Published
- 2025-10-14
- DOI
- https://doi.org/10.55730/1300-0144.6083
- Citations
- 1
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 0.65