Towards a general-purpose foundation model for functional MRI analysis

Abstract

Functional magnetic resonance imaging (fMRI) is crucial for studying brain function and diagnosing neurological disorders. However, existing analysis methods suffer from reproducibility and transferability challenges due to complex preprocessing pipelines and task-specific model designs. Here we introduce the Neuroimaging Foundation Model with Spatial-Temporal Optimized and Representation Modelling (NeuroSTORM), which learns generalizable representations directly from four-dimensional fMRI volumes and enables efficient transfer to diverse downstream applications. Specifically, NeuroSTORM is pretrained on 28.65 million fMRI frames from over 50,000 participants, spanning multiple centres and ages 5-100. It combines an efficient spatiotemporal modelling design and lightweight task adaptation to enable scalable pretraining and fast transfer to downstream applications. We show that NeuroSTORM consistently outperforms existing methods across five downstream tasks, including demographic prediction, phenotype prediction, disease diagnosis, re-identification and state classification. On two multihospital clinical cohorts with 17 diagnoses, NeuroSTORM achieves the best diagnosis performance while remaining predictive of psychological and cognitive phenotypes. These results suggest that NeuroSTORM could become a standardized foundation model for reproducible and transferable fMRI analysis.