A new approach has enabled 71 research centers, including the Institut de recherche sur le cancer de l’Université de Sherbrooke (IRCUS), to rapidly advance knowledge about glioblastoma, a rare and very aggressive form of brain cancer.
Thanks to an innovation combining artificial intelligence and medical imaging, thousands of glioblastoma data could be analyzed without being shared among the 71 participating centers. This important study, in which Professors Martin Vallières, David Fortin and Martin Lepage participated, led to an unprecedented breakthrough in the field: analyzing, in a federated learning context, more than 25,000 magnetic resonance imaging scans, leading to a 33% increase in the tumor border detection rate.