GSS EU Science Funding Gateway

To address these challenges and maximise the opportunity offered by synthetic data, projects funded under this topic should address the following objectives:

• assemble a cross-sectoral public-private consortium including synthetic data experts, public and private data owners, and healthcare solution developers;
• using high-quality public and private datasets, develop / further develop and validate reliable SDG methods for relevant healthcare use cases. The use cases to be explored must be described and justified in the proposal, complement work that is already ongoing, and should:
  • ensure the broad applicability of the SDG methods developed and include data types that are not currently adequately addressed, such as device data, image data, genomic data etc;
  • include methods to generate: a) fully synthetic datasets that do not contain any real data; b) hybrid datasets composed of a combination of data derived from both real and synthetic data; and c) synthetically-augmented datasets.
  • pay particular attention to bias, both in source data and in the SDG methods.
• validate the synthetic data generation methods applied in the project using source data. This should include assessing the risk of re-identification;
• demonstrate the quality and applicability of the synthetic data generated in the project through the development of relevant models;
• encourage the uptake of the results of the project through a strong communication and outreach plan.