IRYCIS advances health data research through cascade funding: PRIV-FHE-VC European project
SECURED - Scaling up secure processing, anonymization and generation of health data for EU cross border collaborative research and innovation - is an initiative funded by the UE whose aim is to launch an EU cross-border health data collaboration ecosystem, so new AI-based data analytics solutions and STEM innovations can be produced.
The consortium tasked with carrying out the project is coordinated by the University of Amsterdam and it represents leading institutions (universities, technological centres, research infrastructures and hospitals) while ensuring a diverse geographical spread across Europe. This 3-year-project with 17 partners (GA n. 101095717) has been granted €7 million, funded within the scope of the Horizon Europe R&D programme, under the call "Scaling up multi-party computation, data anonymization techniques, and synthetic data generation" (HORIZON-HLTH-2022-IND-13-02).
SECURED has specific-dedicated budget to open calls to manage and distribute EU funding to other institutions through the mechanism Financial Support to Third Parties (FSTP). Also known as cascade funding, it is a designed tool by the European Commission to get public funding widely distributed.
Consortia that have received EU grants that allow this FSTP mechanism, are invited to issue open calls and redistribute part of their funding external entities, so they can contribute to specific tasks within the larger project. This approach final aim is to extend the impact of EU programmes. The initiative aims to facilitate real-world testing, expand the usage of SECURED technologies, and generate market potential by addressing critical privacy challenges in healthcare.
At our institution, the proposal is led by Dr. MIGUEL ÁNGEL SICILIA head of Biomedical Data Science and Engineering group (Area 4 - Tools for advanced medicine) and the team carrying out the action is composed by himself and Dr. ALBERTO BALLESTEROS RODRÍGUEZ, collaborator in the same research group.
Named PRIV-FHE-VC - Privacy-preserving predictive inference via FHE and Verifiable Credentials in health contexts - this successful project aims to design and publish a practitioner-oriented benchmark for privacy-preserving inference in medical contexts such as risk assessment and screening. Leveraging Fully Homomorphic Encryption (FHE), representative machine learning models operating on encrypted data are evaluated systematically to derive and report metrics across model families and their inputs. The benchmark will bundle realistic clinical requirements so that practitioners can identify the profiles and resource budgets suited to specific scenarios, with patient information never exposed in the clear.
A final showcase event is expected to be celebrated at the end of the year where all the research teams involved in the project can gather and share their experience during the implementation of the different projects.
"Congratulations to IRYCIS team for its achievement in this open call and for broadening our institution participation in different international programmes and topics as diversification is also a key of success."
