Federated Learning (FL) enables collaborative AI development in health research without sharing sensitive data. By training models locally on decentralized datasets, FL upholds GDPR principles of data minimisation and confidentiality. The TRUMPET EU-funded initiative advances this approach by tackling the technical and regulatory challenges of privacy-preserving computation. This session will showcase how TRUMPET’s federated framework supports secure, large-scale biomedical AI innovation. Organized by: Trumpet Project
What to expect:
Sharing without Sharing: the TRUMPET federated approach for privacy-preserving AI computation
Francesco Ghini – IRST IRCCS
Nuria Barros Reguera GRADIANT
Jaime Loureiro Acuña GRADIANT
Panel discussion and Q&A session
Francesco Ghini – IRST IRCCS
Magdalena Kogut – TIMELEX
Alberto Pedrouzo Ulloa – UVIGO
Zeev Pritzker – Arteevo
Patrick Duflot – CHU Liege