AI-Powered Predictive Models for Stroke Management

CERN’s federated learning platform, CAFEIN, underpins the data ecosystem at the heart of the EU-funded project UMBRELLA, which aims to transform stroke care across Europe over the next five years. The project is driven by  a consortium of over 20 public and private partners co-led by Vall d’Hebron Research Institute (VHIR) and Siemens Healthineers, and kicked off in October 2024.

UMBRELLA will implement a comprehensive approach that addresses gaps along the whole stroke care pathway, from diagnosis and emergency treatment, right through to rehabilitation and the prevention of further strokes. Central to the project is the training and testing of new AI algorithms via a federated learning framework. This approach ensures that the AI models benefit from being trained on multiple datasets, but because the data is not shared or pooled, security and privacy are ensured.

The Innovative Health Initiative (IHI), a partnership between the European Union’s Horizon Europe Research and Innovation programme, industry partners, and industry associations (COCIR, EFPIA, Europa Bío, MedTech Europe, and Vaccines Europe) will fund the project with close to 27 MEUR.

Building on CERN’s expertise, including its success in federated AI modelling as part of the EU TRUSTroke project, the CAFEIN team is poised to handle the increasing complexity and scale of data sets required by UMBRELLA.