Towards Converting Clinical Phrases into SNOMED CT Expressions

Area of Science:

• Medical Informatics
• Natural Language Processing
• Clinical Terminology

Background:

• Automating clinical applications requires converting unstructured clinical text into structured data.
• Standard clinical terminologies like SNOMED CT are crucial for this structured representation.
• Integrating novel clinical phrases into existing terminologies is a key challenge.

Purpose of the Study:

• To develop a method for converting novel clinical phrases into new expressions within SNOMED CT.
• To address the task of identifying relations between clinical phrases and SNOMED CT concepts.

Main Methods:

• A machine learning approach was developed to identify relations between clinical phrases and SNOMED CT concepts.
• The method leverages existing SNOMED CT relations for training and evaluation.

Main Results:

• The proposed machine learning method demonstrates effective performance in identifying relations.
• This facilitates the integration of new clinical information into SNOMED CT.

Conclusions:

• The developed method shows promise for automating the expansion of SNOMED CT with novel clinical phrases.
• This contributes to more comprehensive and accurate clinical data representation.