DPI_for_INAH

Welcome to DPI for INAH's Wiki.

The objective of the DPIforINAH project is to take the INAH project to the next level by implementing a hospital EHR data extractor, based on natural language recognition, to feed the INAH data warehouse in OMOP format. A set of specific analyses will be added to the INAH project in order to medically and technically validate the quality of the transformation of data from the EHR. The project will focus on vaccination, allergy and oncology data.

Key outcomes

Extraction of data (allergies, vaccinations, oncology) from the three DPI systems
Processing of text files using an LLM natural language recognition tool to extract structured data
Depersonalization of data from the three hospitals.
Implementation of a data model for allergies, vaccinations and oncology (SNOMED CT)
Semi-automatic coding of key fields in SNOMED CT
Neutral export of coded data for multiple uses
FHIR-BE labelling for Allergy and Vaccination via FHIR-a-thon
Import into an INAH OMOP database
Medical validation of validated use cases and initial comparison with first line (SUMEHR)

The project was coordinated by Andre Vandenberghe (Andre.VANDENBERGHE@chc.be).