Controlling the vocabulary for anatomy.

Area of Science:

• Anatomy
• Natural Language Processing
• Medical Informatics

Background:

• Natural Language Processing (NLP) systems require comprehensive anatomical references for accurate human anatomy representation.
• Existing electronic anatomical sources are often incomplete, difficult to use, or not tailored for linguistic analysis.
• Discrepancies exist between anatomical and linguistic perspectives on knowledge representation.

Purpose of the Study:

• To review recognized sources of anatomical knowledge for their utility in linguistic analysis.
• To evaluate the potential and limitations of these sources from a linguistic viewpoint.
• To emphasize the significance of the International Federation of Associations of Anatomists (IFAA) work, specifically the Terminologia Anatomica.

Main Methods:

• Literature review of anatomical knowledge sources.
• Analysis of source suitability for natural language processing tasks.
• Comparative evaluation of different anatomical terminologies.

Main Results:

• No single existing electronic anatomical source adequately meets all NLP requirements.
• Sources vary in completeness, usability, and specificity for linguistic applications.
• The Terminologia Anatomica, developed by the IFAA, presents a standardized approach valuable for NLP.

Conclusions:

• A suitable global reference for human anatomy in NLP remains a challenge.
• The Terminologia Anatomica offers a promising, consensus-based resource for improving NLP in anatomical contexts.
• Further development is needed to bridge the gap between anatomical and linguistic data representation.