SEARCH aims to enable secure, large-scale biomedical data analysis by combining synthetic data generation with federated learning to protect privacy while supporting AI-driven research. The project seeks to accelerate personalized medicine through advanced data federation tools, standardized synthetic datasets, and strengthened public-private collaboration across Europe’s digital healthcare ecosystem.

Project description

The objectives of SEARCH are truly ground-breaking, seeking to enable extensive data aggregation and analysis while safeguarding the integrity and privacy of original datasets through synthetically derived proxies. This initiative is designed to address biomedical challenges in Europe and offer translational solutions that will ultimately contribute to the advancement of personalised medicine. Unlike traditional data-sharing platforms that mainly focus on technical obstacles, SEARCH adopts an innovative approach by addressing legal, ownership, and subject privacy concerns. It specifically targets distributed institutional repositories that are hesitant to share multimodal clinical data, overcoming security concerns through a combination of clinical synthetic data proxies and a Federated Learning framework. Until synthetic data proxies gain wider acceptance, this combined strategy aimed at alleviating security concerns, facilitates the scalability required for AI analysis and promotes creative public-private data collaborations.  

SEARCH will offer advanced data federation capabilities, incorporating unique Synthetic Data Generation features to create various data types, including those not comprehensively addressed currently (e.g., wearable device data, image sequences, and genomic data). Through curated access to these novel digital tools, SEARCH will facilitate convenient access for the healthcare industry and research community to address bottlenecks and challenges in the development of novel tools for personalized prevention, diagnosis and treatment based on explainable AI. Moreover, SEARCH will provide agreed-upon gold standard synthetic datasets for evaluating the performance of biomedical AI solutions. SEARCH aims to consolidate European Innovation and Research endeavours by promoting public and private collaborations to unlock the potential for innovation in the digital healthcare sector.