Machine Translation Technology in Health: A Scoping Review

Area of Science:

• Health Informatics
• Medical Translation
• Computational Linguistics

Background:

• Machine Translation (MT) is vital for overcoming language barriers in diverse healthcare settings.
• English dominance in medical research and limited human translators increase MT relevance.
• Recent MT accuracy improvements necessitate re-evaluating its healthcare applications.

Purpose of the Study:

• To assess the appropriateness and limitations of MT in various health contexts.
• To review MT's suitability for medical education, literature translation, and patient-provider communication.
• To identify areas for future research in healthcare MT.

Main Methods:

• Scoping review and meta-analysis of peer-reviewed literature.
• Keyword search across PubMed, PubMed Central, and IEEE Xplore (2018-2023).
• Analysis of 33 selected full-text articles from 2,589 abstracts.

Main Results:

• MT is currently unsuitable for direct patient-provider communication due to clinical risks from inaccuracies.
• MT demonstrates potential for translating medical articles, terminology, and educational materials.
• Significant gaps exist in MT coverage for under-studied languages.

Conclusions:

• MT is not recommended for direct patient interactions in healthcare settings.
• MT shows promise for indirect applications like medical literature and education.
• Further research is needed to enhance MT performance, expand language coverage, and understand current usage patterns.