A Natural Language Processing Tool Offering Data Extraction for COVID-19 Related Information (DECOVRI)

Area of Science:

• Computational linguistics
• Medical informatics
• Public health informatics

Background:

• The COVID-19 pandemic highlighted the need for efficient data extraction from clinical notes.
• Unstructured clinical text poses challenges for large-scale data analysis and research.
• Existing tools may not be specifically tailored for COVID-19 related information extraction.

Purpose of the Study:

• To develop and release a free, open-source natural language processing (NLP) application for extracting COVID-19 related information from clinical text notes.
• To convert unstructured clinical notes into structured data compatible with the OMOP Common Data Model (CDM).
• To support pandemic response efforts through improved data accessibility and usability.

Main Methods:

• Development of a novel NLP application named DECOVRI (Data Extraction for COVID-19 Related Information).
• Focus on extracting key information pertinent to COVID-19 from unstructured clinical text.
• Integration with an OMOP CDM-based ecosystem for structured data storage.

Main Results:

• A functional prototype of the DECOVRI application has been developed.
• The tool is designed to process and structure clinical notes related to COVID-19.
• DECOVRI will be released as a free and open-source tool.

Conclusions:

• DECOVRI offers a valuable solution for transforming unstructured clinical data into a usable format for COVID-19 research.
• The open-source nature of DECOVRI promotes collaboration and wider adoption in the research community.
• This tool is expected to enhance the efficiency of data analysis in response to the COVID-19 pandemic.