SNOMED CT in a language isolate: an algorithm for a semiautomatic translation.

Area of Science:

• Medical Informatics
• Computational Linguistics
• Terminology Management

Background:

• Systematized Nomenclature of Medicine--Clinical Terms (SNOMED CT) is a crucial clinical terminology resource.
• Basque is an official language in the Basque Autonomous Community alongside Spanish.
• Developing resources for less-resourced languages like Basque is essential for equitable access to information.

Purpose of the Study:

• To semi-automatically translate SNOMED CT content into the Basque language.
• To evaluate the feasibility of using Natural Language Processing (NLP) for this translation.
• To support the use of Basque in the biomedical domain.

Main Methods:

• Design and partial implementation of a translation algorithm based on NLP.
• Focus on a four-phase algorithm, with the first two phases completed and evaluated.
• Quantitative evaluation of the implemented translation phases.

Main Results:

• Achieved Basque equivalents for 21.41% of SNOMED CT disorder terms.
• Lower translation rates in other hierarchies: 12.57% for body structure, 8.80% for findings, and 3% for procedures.
• Promising initial results indicate the potential of the developed methods.

Conclusions:

• The translation effort is progressing towards enabling access to multilingual resources in Basque.
• The project aims to strengthen the presence and utilization of the Basque language within the biomedical sector.
• The developed NLP-based approach shows potential for future expansion and improvement.