Innovation Lab : Data heterogeneity and fragmentation
Question:
How can we securely integrate heterogeneous and fragmented health data (e.g., genomic, clinical, behavioral, environmental) into interoperable platforms that enable real-time insights and decision-making?
Proposed solution:
A step toward addressing data heterogeneity and fragmentation in healthcare could involve the development of (LLM-powered) NLP solutions capable of extracting structured information from unstructured patient records. These tools can convert narrative textual information into standardized terminologies such as SNOMED CT.
Once coded, this structured data can be mapped to the OMOP CDM, facilitating semantic interoperability and enabling integration with other clinical, genomic or exposomic datasets, which can be openly, securely and efficiently shared with researchers, healthcare institutions, and industry stakeholders.