FL on IoT For medical applications

The project aims to transform ischemic stroke management by developing a fully patient-centered, privacy-preserving, and AI-driven solution that empowers clinicians, patients, and caregivers across all stages of the stroke care pathway. By deploying advanced federated learning models directly on edge devices, the initiative brings artificial intelligence closer to the patient, enabling real-time monitoring, diagnosis, therapy evaluation, and personalized care, all without the need for intermediate data storage.

This approach aims to bridge the gap between clinical research and real- world patient monitoring, ensuring that sensitive medical data remain securely on local devices while still contributing to the development of robust, trustworthy, and explainable AI models. Building on CERN’s CAFEIN Federated Learning Platform, the project seeks to reshape the way stroke care is delivered by integrating ML algorithms, distributed intelligence, and real-world clinical practices right.

The platform will support longitudinal and continuous patient monitoring, enabling accurate risk estimation, early detection of stroke relapse, optimization of therapeutic strategies, and adaptive rehabilitation strategies. By federating knowledge across multiple healthcare institutions, while preserving strict data privacy, the project will establish a new approach in digital health infrastructures in Europe and beyond fostering more personalized, predictive, and preventive stroke care.