Imagine you need a huge library of patient records to train a medical AI, but hospitals can't give them to you because of privacy laws. This project creates 'fake' but mathematically accurate copies of that data that look and act like the real thing. It's like using a high-quality stunt double in a movie instead of the actual actor to keep the star safe while still getting the scene right.
Medical AI development is stalled because hospitals cannot share sensitive patient data due to GDPR and privacy risks. This creates a data bottleneck that prevents the creation of accurate tools for cancer, stroke, and heart disease.
A federated platform for secure data analysis and a Synthetic Data Assessment and Credibility (SDAC) tool to verify the quality and privacy of synthetic datasets.
If you are a drug discovery firm dealing with fragmented clinical data for cardiovascular or stroke research — this project developed synthetic data proxies that allow you to train AI models without needing direct access to sensitive patient files. This removes the legal bottlenecks of GDPR and data sharing agreements.
If you are an AI developer dealing with a lack of diverse training sets for gynaecological cancer — this project developed a federated platform that enables secure distributed analytics. You can improve your tool's accuracy using data from 14 different countries without the data ever leaving the hospital.
If you are a wearable tech company dealing with the difficulty of getting high-quality genomic or image sequence data — this project developed synthetic generation features for these specific data types. This allows you to build and test personalized prevention tools using gold standard synthetic datasets.
Based on available project data, no specific pricing or cost structures are mentioned; the project is currently in a funded research and development phase.
Yes, the project is designed for scalability through a federated architecture and a consortium of 30 partners across 14 countries to ensure the tools work across different institutional repositories.
Based on available project data, the project is establishing pathways for exploitation and long-term sustainability, but specific licensing terms are not yet detailed.
Yes, it is specifically designed to align with GDPR, the European Health Data Space (EHDS), and the AI Act to ensure regulatory readiness.
The project period runs from 2024-10-01 to 2028-09-30, suggesting a multi-year development and validation timeline.
The project is heavily industry-driven, with 21 industrial partners (70% ratio), including 11 SMEs. This strong commercial presence, combined with 3 universities and 1 research center across 14 countries, indicates a high likelihood of practical commercial application rather than purely academic research.
Contact the Board of the College of the Holy & Undivided Trinity of Queen Elizabeth near Dublin
Contact us to connect with the SEARCH consortium for early access to synthetic health datasets.
