Seed Funding VUB – Technische Universität Dresden – Guy Nagels

Federated learning allows institutions to train AI models on sensitive patient data without transferring the data itself. This initiative will develop a proof-of-concept federated learning framework for neuroimaging biomarkers in multiple sclerosis. By mapping and harmonising data structures and governance at both VUB and TUD, designing a technical and ethical roadmap, and defining clinically meaningful outcomes, the project aims to build a sustainable multi-centre research platform that respects legal and ethical constraints while enabling collaborative AI model training.