Data Sciences Speaker Series/Temerty Centre Speaker Series: Dr. Sabine Österle

Thanks to everyone who attended this insightful talk. We'll post the video recording to our Past Events page as soon as it becomes available.

____________________

Presentation Title

Swiss Personalized Health Network – A national, graph-based framework for the semantic representation of FAIR health data

Abstract

The Swiss Personalized Health Network (SPHN) has developed a national framework and tool stack for the semantic representation of health data in a knowledge graph. This framework has been implemented in all Swiss university hospitals to facilitate the sharing and integration of various types of health-related data from different sources. The goal is to advance medical research through the availability of health data prepared in accordance to the FAIR (Findable, Accessible, Interoperable, Reusable) principles.

To enable researchers to build medical knowledge graphs in a simplified way, SPHN provides services and a tool stack for the easy design, generation, and validation of RDF graph data from multiple sources. At its core, a common SPHN data schema in RDF provides around 90 basic as well as more complex compositions of SPHN concepts. These concepts can be used directly by the projects or integrated into new concept compositions. New concepts can be easily designed in an Excel spreadsheet and are translated by the SPHN Schema Forge web service into an RDF schema. Furthermore, Schema Forge provides researchers with design validation rules and queries for basic statistics of the data, all in less than 5 minutes. The DCC Terminology Service provides external national and international standards like SNOMED CT, LOINC, CHOP, and ATC in a SPHN compatible form, enabling seamless integration. Lastly, the SPHN Connector is a tool designed to facilitate the generation of de-identified graph data according to a well-defined RDF schema.

The standardization of the semantic representation of health data using graph-based technology created within SPHN enables efficient and accurate combination of data from different sources. It allows the use and integration of additional knowledge from terminologies and classifications so that a comprehensive overview of the patient’s health status along the treatment pathway in the hospital is possible. The infrastructure components of the framework are designed to facilitate scaling to include additional healthcare facilities and other data providers.

Learning Objectives

• To understand the Swiss Personalized Health Network (SPHN) framework for the semantic representation of health data in a knowledge graph.

• To enable researchers to build medical knowledge graphs in a simplified way: a tool stack for the easy design, generation, and validation of RDF graph data from multiple sources, and to explore the potential to apply them to your project.

• To explore the challenges and limitations associated with using clinical routine data for research and knowledge graph development, including issues related to data quality, privacy concerns, data silos, and interoperability hurdles.

• To learn about the limitations of existing medical standards and SPHN's approaches to making such vocabularies more FAIR (Findable, Accessible, Interoperable, Reusable principles) and more usable.

About the speaker

Sabine Österle is a team lead for data interoperability in the Personalized Health Informatics (PHI) group of the SIB Swiss Institute of Bioinformatics. The PHI group manages the Swiss Personalized Health Network (SPHN) Data Coordination Centre (DCC) as well as the BioMedIT project. SPHN is a national initiative with the goal of developing, implementing and validating a coordinated data infrastructure, in order to make health-relevant data interoperable and shareable for research in Switzerland. An integral part of SPHN is the BioMedIT project: a nationally trusted IT environment for sensitive health data for research.

Sabine Österle received her BSc and MSc degrees from ETH Zurich in Interdisciplinary Science in 2010 and 2012, respectively. She then pursued a PhD degree in Synthetic Biology at the Department of Biosystems Science and Engineering of ETH Zurich, after which she decided to focus on personalized health. In her current role at SIB, she leads the data interoperability team of PHI which coordinately develops the national semantic interoperability framework and related tool stack. Sabine and her team also provide consultancy on semantic interoperability within the SPHN context.