SYNTHEMA
banner
synthemaeu.bsky.social
SYNTHEMA
@synthemaeu.bsky.social
A cross-border hub to develop & validate #AI techniques for anonymization & synthetic data generation in rare hematological diseases 🔬🩸🇪🇺 #HorizonEU project

👉 https://synthema.eu/
How similar is synthetic tabular data to the real thing?

This new approach uses KL and JS divergences to evaluate full joint distributions — not just feature by feature. A more reliable way to validate synthetic datasets in practice.

👉 synthema.eu/publications/

#SyntheticData #DataValidation #AI
December 1, 2025 at 10:52 AM
Over 450 haematological diseases affect blood cells, the lymphatic system & coagulation, posing major challenges to public #HealthSystems.
#SYNTHEMA is developing a secure platform for synthetic #ClinicalData to enable GDPR-compliant research in rare #HaematologicalDiseases.

👉 synthema.eu
November 25, 2025 at 9:27 AM
Can synthetic tabular data be more accurate?

This new model by researchers at Universidad Politécnica de Madrid combines a Variational Autoencoder with a Gaussian Mixture Model. Results outperform #CTGAN and #TVAE, especially on medical datasets.

👉 synthema.eu/publications/

#SYNTHEMA
November 20, 2025 at 9:20 AM
The paper “Distributing trust: A P2P SMPC protocol for secure federated learning” by Petros Demetrakopoulos, Themistoklis Anagnostopoulos and Sofia Tsekeridou (Netcompany RID, #SYNTHEMA partner) will be presented at #FLLM2025 on 27 Nov at 15:30.

synthema.eu
November 17, 2025 at 9:48 AM
#HealthData4EU Cluster is at #EBDVF25 this week. SYNTHEMA isn’t on site, but we’re glad to support tomorrow’s session in Copenhagen. If you're there on 14 Nov, join “Ecosystems of Innovation in Healthcare” from 09:00–10:30, Room 2.B — a chance to hear how our partners advance ethical health data use
November 13, 2025 at 9:16 AM
How well does synthetic data generation work when spread across multiple hospitals? This study on #AML shows that federated learning protects privacy, though fidelity takes a hit. A key step in real-world AI for rare diseases.

👉 synthema.eu/publications

@vicomtech.bsky.social

#SYNTHEMA
November 10, 2025 at 9:49 AM
Two productive days in San Sebastian for the #SYNTHEMA consortium have come to an end. Lively discussions, clear goals for the final year: a recap publication on synthetic data in haematology, joint training with @erneurobloodnet.bsky.social, new papers, and the SYNTHEMA platform launch

synthema.eu
November 6, 2025 at 10:00 AM
The SYNTHEMA consortium is meeting in San Sebastian, Spain, to mark the project’s third anniversary and kick off discussions for the final year ahead. Together, we’re committed to delivering results in secure, AI-powered health data for rare hematological diseases.

👉 synthema.eu
November 4, 2025 at 9:12 AM
What can generative models reveal about the real data they learn from? This study by @upm.es researchers explores membership inference attacks in generative models, introducing new theory and showing how differential privacy fits in. A critical read for privacy-aware AI: synthema.eu/publications/
October 30, 2025 at 8:39 AM
Pushing the boundaries of AI-based research in rare haematological diseases.

#SYNTHEMA develops privacy-preserving AI and synthetic data to create virtual patients that improve diagnosis and treatment in rare diseases.

👉 synthema.eu

#AIinHealthcare #HorizonEurope
October 28, 2025 at 10:25 AM
The #HealthData4EU Cluster, which SYNTHEMA forms part of, will host the session “Ecosystems of Innovation in Healthcare: Ensuring Trust, Quality & Ethics in Data Use” at #EBDVF2025 in Copenhagen on 12–14 Nov

Event: european-big-data-value-forum.eu/2025-edition...
Cluster: synthema.eu/healthdata4eu/
October 20, 2025 at 8:50 AM
Most synthetic data models assume data abundance. This study shows how transfer learning and meta-learning improve synthetic data generation for cancer research in low-data scenarios, where real clinical constraints apply.

👉 synthema.eu/publications/

#SYNTHEMA #SyntheticData #AI
October 16, 2025 at 8:21 AM
Our partner i~HD is hosting a special 10th anniversary edition of its Annual Conference: “A Decade of Joining the Dots – Towards Bridging Data Spaces”

📍 3–4 Dec 2025 | Ghent, Belgium

Topics: EHDS, AI in healthcare, data quality, privacy & governance.

👉 www.i-hd.eu/ac2025
October 13, 2025 at 9:52 AM
Most synthetic data models assume data abundance. This study shows how transfer learning and meta-learning improve synthetic data generation for cancer research in low-data scenarios, where real clinical constraints apply.

👉 synthema.eu/publications/

#SYNTHEMA #CancerResearch
October 6, 2025 at 8:25 AM
In #SYNTHEMA, data collection is central to building GDPR-compliant, AI-ready datasets for Sickle Cell Disease and Acute Myeloid Leukaemia. Harmonised across countries, the data supports synthetic generation, federated learning, and future rare disease research.

👉 synthema.eu/2025/09/23/d...
October 2, 2025 at 7:39 AM
Data scarcity and fragmentation limit synthetic data in healthcare. This study introduces Synthetic Data Sharing, a federated learning method that improves representativeness and reduces disparities between data-rich and data-poor institutions.

👉 synthema.eu/publications/

#SYNTHEMA
September 29, 2025 at 7:52 AM
Good data is not just collected, it is built with care. In #SYNTHEMA, clinical teams across Europe are creating high-quality, privacy-compliant datasets to power synthetic data and AI research in rare blood disorders.

👉Read more: synthema.eu/2025/09/23/d...
September 25, 2025 at 8:08 AM
Cyber-physical systems need smarter tools for risk assessment. This study introduces an automated, simulation-based method aligned with ISO 27005. Validated through a real attack case and implemented in Spyderisk.

Read more 👉 synthema.eu/publications/
September 22, 2025 at 8:35 AM
📰 The #BDVA Healthcare Task Force has published its white paper on Synthetic Data in Healthcare.

Proud to contribute alongside AISym4Med & SECURED, our sister projects in the #HEALTHDATA4EU cluster.

Read more 👉 bdva.eu/news/synthet...

#SyntheticData #SYNTHEMA
September 19, 2025 at 8:42 AM
#SYNTHEMA is featured in The Innovation Platform (Issue 23), with an article on how synthetic data and federated AI are transforming research in rare blood disorders while keeping patient privacy at the core: www.innovationnewsnetwork.com/publication/...

#SyntheticData #RareDiseases
September 15, 2025 at 9:46 AM
The #SYNTHEMA platform enables synthetic health data without moving sensitive patient info.
✔️ Data stays local
✔️ Federated learning across centres
✔️ FAIR-based quality & compliance

Tested in #SCD & #AML use cases. More info: synthema.eu
September 11, 2025 at 9:16 AM
#SYNTHEMA is building the foundations for a future where synthetic, privacy-preserving data supports progress in rare diseases.

Beyond its two use cases, its flexible design can benefit other rare and underrepresented conditions.

More 👉 synthema.eu
September 9, 2025 at 8:37 AM
Estimating treatment effects in healthcare is complex when data is decentralized and privacy is critical. This study applies federated learning to generative causal models for privacy-preserving ITE estimation across fragmented datasets.

👉 synthema.eu/publications/

#SYNTHEMA
September 2, 2025 at 8:05 AM
Local data stays safe. Global research moves forward. With SYNTHEMA’s federated learning, hospitals never share patient data, but together they unlock new possibilities for rare disease research.

👉 synthema.eu
August 25, 2025 at 8:52 AM
Sickle Cell Disease is a rare, life-threatening blood disorder with limited treatment options. In #SYNTHEMA’s Use Case 2, AI creates synthetic data from 1,000+ patients to study genetic factors & predict brain vascular events, while protecting privacy.

👉 synthema.eu/use-cases/
August 19, 2025 at 7:58 AM