Voir en


Webinar on Federated Learning on 28 September

Find out more about federated learning and how it can be used in healthcare.

Webinar poster


A federated learning (FL) platform, called CAFEIN*, based on artificial-intelligence (AI) algorithms, was developed at CERN in order to ensure immense precision in the operation of the complex accelerator chain.

The key advantages of FL are:

  • Privacy protection: Raw data never leaves the device, which protects user privacy and sensitive information.
  • Efficiency: Federated learning can be more efficient than traditional centralised training, especially for large datasets or when dealing with data stored on edge devices.
  • Decentralisation: It's well-suited to scenarios where data is distributed across many devices or locations, such as IoT devices, healthcare institutions and more.
  • Adaptation: Federated learning can be used to adapt models to individual user preferences or local conditions while maintaining a global model’s performance.

These special features make it useful in applications where data privacy is a top priority, such as in mobile devices, healthcare and more.

You are warmly invited to join this online webinar to find out more about CAFEIN and its applications in various domains, with a particular emphasis on healthcare through a new EU project (TRUSTroke)**. The speakers include:

  • Stefano Savazzi (Consiglio Nazionale delle Ricerche), who will talk about FL theories and applications;
  • Michele Carminati and Alessandro Redondi (Politecnico di Milano), who will cover the federated network, its security and its privacy;
  • Luigi Serio and Diogo Reis Santos (CERN), who will explain and demonstrate the CAFEIN FL operating platform.

More information: https://indico.cern.ch/e/trustrokewebinar


* The Computer-Aided deFEcts detection, Identification and classificatioN (CAFEIN) project has received support from the CERN Budget for knowledge transfer to medical applications through a grant awarded in 2019

**The TRUSTroke project is funded by the European Union in the call HORIZON-HLTH-2022-STAYHLTH-01-two-stage, under grant agreement No-101080564