PEPR TREASURE · Work Package 1
Augmented data
A multidimensional, interoperable transplantation dataset
WP1 is progressively building a large data repository to improve the care of transplant patients. This collection will form the core of WP6 — the very first AI-based transplant clinic.
- Interoperability
- Predictive AI
- Governance
- Data security
Axes and objectives
- 1
Multi-source collection and integration
Bring together and aggregate highly diverse data — electronic health records, genomics, imaging, patient outcomes, biobanks — as well as all the data produced by the other work packages (WP2 to WP5).
- 2
Structuring and technical architecture
Distinguish structured from unstructured data (imaging), automate collection through ETL and APIs, and build a single, patient-centred architecture — deterministic (NIR) or probabilistic linkage, cloud and on-site storage, resilient backups.
- 3
Interoperability and security
Rely on the HL7 FHIR and DICOM standards, create federated query tools (analyse sources without moving the data), and guarantee confidentiality and security in line with the French Ministry of Health’s best practices.
- 4
Governance and proofs of concept
Build a robust governance framework with institutional and healthcare stakeholders (Biomedicine Agency, hospitals, Health Data Hub, CNIL, CNAM, patient associations…) and start with proofs of concept to validate the architecture and the linkage of existing databases.
- 5
Expected impact
Very high-quality data to develop powerful predictive AI algorithms and advanced analytics: personalise interventions, optimise organ allocation, improve diagnosis and forecasting, and position France as a world leader in transplantation research.