Fed-BioMed
Open-Source Platform for Secure Collaborative Health Data Programs
Open, transparent and trusted federated technologies for real-world healthcare applications
Medical data cannot move.
But medical research must.
Patient data cannot be freely shared
Healthcare data providers must protect privacy and trust
A single institution does not have enough data
Research requires collaboration across data providers
Healthcare data providers operate under strict responsibility.
- Patient data must remain protected at all times.
- Legal and ethical obligations limit data sharing.
- IT systems and infrastructures vary across institutions.
- Research collaboration requires clear governance and accountability.
At the same time, high-quality medical research increasingly depends on data from multiple institutions.
Balancing collaboration and responsibility remains a central challenge.
Move computation, not the data.
Enable secure collaboration across institutions, without compromising data privacy.
Connects distributed institutions in a secure federated network.
Ensures secure and governed access to federated resources, with permissions controlled by participating institutions.
Reveals what data is available across sites through shared metadata.
Helps explore and understand distributed data without moving it.
Trains models collaboratively while keeping data within each institution.
Governance
Healthcare data providers approve and control every training activity
Explore Governance ➤Accessible user experience
Designed for real-world healthcare teams
Explore Accessible user experience ➤Easy deployment
Deploy and operate a federated node quickly, with minimal configuration
Explore Easy deployment ➤Who it is for?
Healthcare Data Providers
- Keep full control of patient data
- Approve and monitor research activities
- Collaborate without centralizing data
Researchers & Data Scientists
- Work with multi-site data
- Run collaborative experiments
- Access real-world clinical environments
Developers & Contributors
- Contribute to an open-source framework
- Improve security and usability
- Build for real healthcare constraints
Used in real healthcare environments.
- Multi-center hospital collaborations
- European research projects
- Real clinical deployments
- Ongoing development with healthcare partners
We are proud of our works
We want to bring solutions to improve the healthcare world.
20+
Collaborative research across institutions
100%
Data stays at source
"Fed-BioMed enables hospital partners from Unicancer to develop robust, decentralized, large-scale AI models in oncology without compromising patient confidentiality, while keeping hospitals actively involved throughout the entire process. These projects are generating strong enthusiasm across the medical community."
Olivier Humbert
Professor in Nuclear Medicine, Centre Antoine Lacassagne & Researcher, Institut de Biologie Valrose
"Fed-BioMed is a great and flexible framework with comprehensive documentation. I was able to seamlessly transition from an initial exploratory phase to a real-world federated deployment. The development team is very friendly and provided great support whenever I was stuck."
Michelle Wang
PhD Candidate, McGill University
"Fed-BioMed allows us to collaborate securely on sensitive data, accelerating clinical research while maintaining full control and privacy."
Renaud Schiappa
Chief Data Officer, Centre Antoine Lacassagne
"Fed-BioMed’s versatile interface allowed me to quickly convert an existing machine learning model into a federated workflow and efficiently adapt it to multi-hospital data."
Romain Andres
AI Engineer, François Baclesse Cancer Center & AImaging Group (ISTCT Unit)
Ready to get started?
Fed-BioMed is fully open-source and available today
pip install fedbiomed[node, gui]
pip install fedbiomed[researcher]