SYNTHEMA is a Horizon Europe research and innovation action that aims to establish a cross-border health data hub for rare haematological diseases (RHDs).

Haematological diseases are highly diversified, with oncological and non-oncological subcategories. The scarcity and fragmentation of patient data across scattered transnational repositories hinder effective health planning and make difficult to engage in basic and clinical research. SYNTHEMA aims to tackle this challenge by establishing a research platform  connecting clinical centres of excellence in the research and care of RHDs, technical  research centres, industries and SMEs, to advance translational and clinical research by generating and validating anonymised and synthetic data in RHDs.

To ensure data security and patients’ privacy, the project will make use of a federated learning (FL) infrastructure, privacy preserving by design, that will allow to iteratively train, refine and validate AI algorithms at hospital premises, without sharing of data outside local repositories. Also, it will equip it with secure multi-party computation (SMPC) protocols and differential privacy (DP), allowing distributed computation of mathematical functions between centres without revealing the underlying data.

SYNTHEMA will focus on two representative RHD use cases: sickle-cell disease (SCD), for non-oncological haematologic diseases, and acute myeloid leukaemia (AML) for haematologic cancers.

To make the project vision a reality, the research is articulated into five strategic objectives, as seen below:

  1. Provide novel methods and capabilities to generate synthetic multimodal clinical, omics, and imaging data for RHDs with a validated clinical result.
  2. Develop de-identification and anonymisation pipelines at the service of clinical research and care.
  3. Consolidate and scale-up the use of FL applications, SMPC and DP solutions for privacy-preserving local algorithm training and global model aggregation.
  4. Ensure ethical and GDPR compliance in anonymised and synthetic data-driven research in RHDs.
  5. Ensure wide uptake and scalability of the developed methodologies and tools through effective stakeholder engagement, dissemination and open science practices.

SYNTHEMA will contribute to existing data registries such as the European Rare Blood Disorders Platform (ENROL), the European Platform on Rare Disease Registration (EU RD Platform), and the European Rare Disease Registry Infrastructure (ERDRI) with data standards, pipelines, and shareable data assets, and support their long-term sustainability.

Visit their website and stay tuned to their social media channels, Twitter and LinkedIn!