Exploring Turkish equivalents of terms for musculoskeletal radiology: insights for a standardized terminology

Aims: This study aimed to provide an analysis of Turkish equivalents of English terms for musculoskeletal radiology. Methods: The present study focuses on a global endorsement of English terms in musculoskeletal radiology, and explores how their Turkish equivalents are used in reference books (Turkish translation of the books, Diagnostic Imaging: Musculoskeletal: Trauma and Diagnostic Imaging: Musculoskeletal: Non-Traumatic Disease). Furthermore, the study attempts to provide a picture of how AI-based tools (i.e. neural machine translation tools such as DeepL, Google Translate and an AI Chatbot, ChatGPT) vary in the translation of these terms. Results: The study found that the most common translation strategies for musculoskeletal radiology terms were borrowing and literal translation, with several combined strategies used for complex terms. AI-based tools like DeepL, Google Translate, and ChatGPT showed a high similarity to human translations, but differences were observed in word choice, strategy use, and orthographic variations. These differences, though minor, highlight the challenges of achieving consistency and accuracy in AI-generated medical translations. Conclusion: The present study provides a list of Turkish equivalents for musculoskeletal terminology in English, and presents an analysis of translations by radiology specialists and AI-based tools. Careful evaluation of AI translations is essential to ensure accuracy and consistency in the translation of medical terminology, particularly in subspecialities such as musculoskeletal radiology.