Improving Health Research in EU through FAIR Data

The FAIR (Findable, Accessible, Interoperable, Reusable) guiding principles were conceptualized to pave the way to machine accessible and actionable data interoperability. There was a growing interest to and application of the FAIR principles in different domains. Health data reseachers have been the leading community to benefit from the FAIR principles for the sake of health data sharing to address different challenges such as better healthcare management and secondary use of healthcare data for biomedical and clinical research.

With ~3 million € budget, the FAIR4Health project aimed to encourage the health research community to FAIRify, share and reuse their datasets. The project facilitated this by providing a rich set of software systems and implementing impressive demonstrations within real-world settings to show the potential impact that such strategy can have on health outcomes.

I can honestly say that this was the R&D project that I worked a lot, I learned a lot and I enjoyed the most. I worked with the researchers of 17 different organizations from 11 different countries. I had the chance of leading the technical development of the FAIR4Health project under the scientific coordination of Prof. Carlos L. Parra-Calderón and his team from the Technological Innovation Unit at Virgen del Rocio University Hospital as part of the Andalusian Health Service in Spain.

With the international team that I was leading, we approached the FAIR4Health software systems from two different perspectives inline with the project objectives. On one hand, we designed and developed a FAIRification toolset so that existing healthcare and health research datasets can be transformed into FAIR datasets utilizing a well known and popular standard, namely HL7 FHIR. On the other hand, we built a federated machine learning architecture so that the FAIR datasets of different institutions can be used to create machine learning models without moving data out of its owner premises. I carried out the following tasks in the context of the project:

  • I designed the common data model for the FAIR4Health Project utilizing HL7 FHIR and FHIR Profiling. This model can be found on GitHub.
  • I have been the lead architect and developer of the Data Curation Tool and the Data Privacy Tool.
  • I did lead an international team of software engineers, machine learning experts and data scientists to develop the Federated Machine Learning System.