Case 2: Sharing Synthetic Data Using AI In the healthcare field, sharing information is key to advancing research and developing new medical tools, but doing so without compromising patient privacy is a major challenge.Synthetic data, artificially generated using artificial intelligence to mimic real data, offers an innovative solution. This approach allows researchers and healthcare professionals to work with information that reproduces the characteristics of real clinical data without revealing personal or sensitive information.In this use case, tools and methods are developed to create synthetic data from medical records and biomarkers, enabling safe and ethical testing of AI algorithms, analyses, and clinical trial simulations. This approach fosters collaboration between institutions, reduces development costs and time, and accelerates scientific progress while respecting patient privacy. Aims Analyze, share, and integrate data among the different entities that provide it—HUVH, VHIR, and VHIO—complying with the established data governance. Model and harmonize the data according to the OMOP common data model. Generate synthetic data while maintaining the OMOP common data model, promoting the ability to easily and scalably incorporate other data sources that follow the same model. Ensure the quality of the generated data so that it can be shared within the data space. Publish them in the data space component catalog so that potential consumers can explore and use them. Data consumers Researchers and healthcare professionals: professionals from the Vall d’Hebron Campus or external organizations conducting studies or innovation projects without needing access to sensitive patient data. Pharmaceutical industry and regulatory agencies: companies that use high-quality synthetic data derived from a real hospital environment, which can be employed in non-randomized clinical trial testing of promising drugs, especially in the context of precision medicine (treatment selection based on biomarkers and genetic data). Companies and developers: companies or developers looking to validate their information systems or medical devices and train AI/ML models without compromising patient data. Health authorities: managers who use the data to make decisions, model policies, or run scenario simulations, such as assessing the impact of approving a new diagnostic or therapeutic technology. Case 1: Health Outcomes and European Health Data Space This case demonstrates how patient-provided data can be used to deliver more personalized and effective care. The goal is to make better decisions, tailor care to each individual’s needs, and share information securely between hospitals and healthcare centers. More information Case 3: Data Space for Federated Prediction of Acute Heart Failure Risk This case demonstrates how data from multiple hospitals can be used securely and collaboratively to create a model that helps predict complications and readmissions in heart failure patients, improving decision-making and patient care without compromising privacy. More information