FL4Health

FL4Health is an open-source Federated Learning (FL) engine: It streamlines secure, cross-silo AI model development for healthcare, prioritizing patient data privacy.

This is FL4Health: a flexible, open-source Federated Learning engine built by the Vector Institute’s AI Engineering team. It extends the foundational Flower library, providing composable modules for distributed optimization and advanced privacy-preserving methods. The platform enables institutions, like those in the Gemini Consortium, to collaboratively train robust machine learning models on diverse datasets without sharing sensitive patient data. FL4Health’s core value is accelerating state-of-the-art FL research and deployment in clinical settings, ensuring models improve continuously while maintaining stringent data security standards.

https://vectorinstitute.github.io/FL4Health/