The Role of ChatGPT-4o in Radiology Assistant Training in Meniscal Tears and Rotator Cuff Pathologies
Yazarlar (4)
Doç. Dr. Hüseyin Alper KIZILOĞLU Tokat Gaziosmanpaşa Üniversitesi, Türkiye
Prof. Dr. Murat BEYHAN Tokat Gaziosmanpaşa Üniversitesi, Türkiye
Suat Kubat
Türkiye
Türkiye
Dr. Öğr. Üyesi Yaşar BİRİŞİK Tokat Gaziosmanpaşa Üniversitesi, Türkiye
| Makale Türü | Özgün Makale (Diğer hakemli uluslarası dergilerde yayınlanan tam makale) | ||
| Dergi Adı | International Journal of Tokat Medical Sciences | ||
| Dergi ISSN | 3061-9912 | ||
| Dergi Tarandığı Indeksler | EBSCO | ||
| Makale Dili | İngilizce | Basım Tarihi | 12-2025 |
| Cilt / Sayı / Sayfa | 17 / 3 / 124–134 | DOI | – |
| Makale Linki | https://dergipark.org.tr/tr/pub/ijtms/article/1758417 | ||
| Özet |
| Objective ChatGPT, which is widely used in daily practice with the developing artificial intelligence technology, is based on natural language processing and deep learning techniques. The aim of this study was to demonstrate the accuracy and contribution of ChatGPT as a supportive training model for radiology assistants. Materials and Methods In this study prepared as a nonhuman study, some questions were prepared from ‘radiologyassistant’ and ‘radiopedia’ education sites and the questions were asked to 5 assistants separately. The same questions were also asked to ChatGPT 4o. The correct answers to the questions were compared. The agreement of the assistants and the ChatGPT 4o model was analysed. Results ChatGPT-4o model gave correct answers to 41 questions out of 48 questions in total. In total, the accuracy rate is 85.4%. In total questions, ChatGPT-4o model was shown to contribute to radiology education for only 1 assistant. There was no significant difference between other assistants and ChatGPT-4o model in answering the questions correctly. In the evaluation of 5 assistants among themselves, a statistically significant difference was found in the answers given to the questions (p=0,03). Conclusion Although the ChatGPT-4o model is a resource to be used in radiology education, its contribution to assistant education is limited. The software needs to be improved. |
| Anahtar Kelimeler |
