Linked data in SPHN - A framework to semantically express and enrich health-related data
Presenters: Dr. Sabine Österle, Team Lead Data Interoperability, and Dr. Vasundra Touré, Scientific Coordinator, Kristin Gnodtke, Senior Clinical Data Specialist, Personalized Health Informatics Group, SIB Swiss Institute of Bioinformatics
About this Webinar:
Today, healthcare information is collected in very diverse systems. Often, ad hoc databases or data models are created for a specific medical use case. When dealing with heterogeneous and sensitive health-related data in a research setting, one of the biggest challenge is to bring the data together and facilitating data analysis. The Swiss Personalized Health Network (SPHN) provides a semantic framework to foster interoperability of health-related data across the fragmented Swiss healthcare systems. During the past two years, a Semantic Web based FAIR (Findable, Accessible, Interoperable and Reusable) data framework has been developed to define, represent and store clinical data using common semantics and formats. Biomedical data from any institution, implementation or platform can therefore be expressed in this framework and semantically annotated following the same rules. The encoded data is presented in a graph model, using the Resource Description Framework (RDF) format in our case, which enables the exploration of data not only within but also between institutions. Futhermore, the flexibility offered by the graph enables the reuse of existing standard terminologies of interest such as ICD-10, LOINC or SNOMED-CT and is extendable for the unforeseen. This allows the researcher to link the knowledge of external terminologies to their data for even more extended exploration using, for instance, reasoning possibilities thanks to the semantic hierarchies povided by these terminologies.
Date: 02 June 2021
Time: 11:00 - 12:00
Register: click here.
If you would like to be added to the distribution/invitation list please complete the contact form on the DCC Seminar and Training page (click here).