Fostering Trust in AI driven Healthcare: SecUre and uNbiased knowleDge guided gEneRative AI – THUNDER

Machine learning (ML) has a future in the healthcare sector, but limited data setsand privacy regulations constrain progress. Generative AI can produce synthetic data, but trust issues with fairness and privacy hinder its use. With the support of the Marie Skłodowska-Curie Actions programme, the THUNDER project aims to address gaps in trustworthy generative models and machine learning methods for healthcare, particularly around standardised evaluation frameworks and trust issues. Its main goal is to develop a comprehensive framework for responsible generative AI in healthcare, with an emphasis on defining evaluation metrics, creating knowledge-guided generative models, and designing interpretable learning models. The project will particularly target sepsis, a global health priority designated by the World Health Organization (WHO).