MEDICAL TRANSLATION IN THE AGE OF NEURAL MACHINE TRANSLATION: STANDARDS, ASSESSMENT MODELS, AND ROMANIAN PERSPECTIVES
MEDICAL TRANSLATION IN THE AGE OF NEURAL MACHINE TRANSLATION: STANDARDS, ASSESSMENT MODELS, AND ROMANIAN PERSPECTIVES

Summary/Abstract: This article explores the evolving role of machine translation (MT), post editing (PE), and translation quality assessment (TQA) in the context of medical translation, with a particular focus on the English–Romanian language pair. Methodologically, the paper offers a narrative, critical review and comparative synthesis of standards and TQA models applied to medical translation, with emphasis on the Romanian context. It examines the historical development of MT technologies, from rule based systems to neural MT (NMT), and assesses their applicability to medical texts. The study highlights the importance of post editing standards—especially ISO 18587—and discusses the cognitive and procedural demands of PE in high stakes domains. A comparative analysis of key TQA models—including those proposed by House, Waddington, the American Translators Association (ATA), MQM, and DQF—reveals their strengths and limitations in evaluating medical translations. The article also identifies significant deficits in Romanian translation research and practice, particularly in doctoral inquiry, pedagogical integration, and professional standardization. In response, it proposes directions for research, teaching, and professional development that aim to consolidate the role of PE and TQA in Romanian medical translation. By bridging theoretical models with practical applications, the study contributes to the advancement of Translation Studies and advocates for a more coherent and responsive framework for ensuring the quality and reliability of medical translations.

• Details
• Contents