Federated learning for MRI-based BrainAGE: A multicenter study on post-stroke functional outcome prediction.
Authors
Affiliations (9)
Affiliations (9)
- Univ. Lille, CNRS, Inserm, CHU Lille, Institut Pasteur de Lille, US 41 - UAR 2014 - PLBS, F-59000 Lille, France. Electronic address: [email protected].
- Univ. Lille, Inria, F-59000 Lille, France.
- Univ. Montpellier, Inria, F-34090 Montpellier, France.
- Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, USA.
- CHU Lille, Département de Neurologie vasculaire, F-59037 Lille, France.
- CHU Lille, Département de Neurologie vasculaire, F-59037 Lille, France; Univ. Lille, Inserm, CHU Lille, U1172 - Lille Neurosciences & Cognition, F-59037 Lille, France.
- CHU Lille, Département de Neuroradiologie, F-59037 Lille, France.
- Univ. Lille, CNRS, Inserm, CHU Lille, Institut Pasteur de Lille, US 41 - UAR 2014 - PLBS, F-59000 Lille, France; Univ. Lille, Inserm, CHU Lille, U1172 - Lille Neurosciences & Cognition, F-59037 Lille, France; CHU Lille, Département de Neuroradiologie, F-59037 Lille, France.
- Univ. Lille, CNRS, Inserm, CHU Lille, Institut Pasteur de Lille, US 41 - UAR 2014 - PLBS, F-59000 Lille, France; Univ. Lille, Inserm, CHU Lille, U1172 - Lille Neurosciences & Cognition, F-59037 Lille, France; CHU Lille, Département de Médecine Nucléaire, F-59037 Lille, France.
Abstract
Brain-predicted age difference (BrainAGE) is a neuroimaging biomarker reflecting brain health, with potential implications for post-stroke recovery. However, training robust BrainAGE models requires large, diverse datasets, often restricted by privacy and regulatory concerns. This study evaluates the performance of federated learning (FL) for BrainAGE estimation in ischemic stroke patients treated with mechanical thrombectomy, and investigates its association with clinical phenotypes and functional outcome. We used pre-treatment FLAIR brain images from 1674 stroke patients across 16 hospital centers. We implemented standard machine learning and deep learning models for BrainAGE estimates under three data management strategies: centralized learning (pooled data), FL (local training at each site), and single-site learning. We reported prediction errors and examined associations between BrainAGE and vascular risk factors (e.g., diabetes mellitus, hypertension, smoking), as well as functional outcome at three months post-stroke. Logistic regression evaluated BrainAGE's predictive value for this outcome, adjusting for age, sex, vascular risk factors, stroke severity, time between MRI and arterial puncture, prior intravenous thrombolysis, and recanalisation outcome. While centralized learning yielded the most accurate predictions, FL consistently outperformed single-site models. BrainAGE was significantly higher in patients with diabetes mellitus across all models. Comparisons between patients with good and poor functional outcome, and multivariate predictions of these outcome showed the significance of the association between BrainAGE and post-stroke recovery. FL enables accurate age predictions without data centralization. The strong association between BrainAGE, vascular risk factors, and post-stroke recovery highlights its potential for personalized prognostic modeling in stroke care.