The Reorganization of Cerebellar Functional Network Topology in Post-stroke Aphasia: A Resting-State fMRI Study

Post-stroke aphasia (PSA) is a prevalent complication of left-hemispheric stroke, yet cerebellar contributions to language and cognition remain insufficiently characterized at the network level. We investigated cerebellar functional network topology in PSA and its associations with language and cognitive performance. Seventy-three right-handed PSA patients and 75 matched healthy controls underwent 3T resting-state functional magnetic resonance imaging (rs-fMRI). The cerebellum was parcellated using the Seitzman-27 atlas, which assigns cerebellar ROIs to canonical large-scale networks, allowing network-specific characterization and interpretable comparisons across functional systems. Global and nodal topological metrics and ROI-to-ROI functional connectivity were quantified, and their relationships with language and non-language cognition were assessed. Small-world organization was preserved in both groups, but PSA patients showed significant reductions in the small-world index, clustering coefficient, and local efficiency, indicating impaired network segregation and local processing. Nodal analysis revealed decreased degree centrality in the frontoparietal network and increased degree centrality in the dorsal somatomotor network. Functional connectivity was reduced within the default mode and frontoparietal networks, while connectivity between frontoparietal and dorsal somatomotor regions increased (FDR-corrected). Lower clustering coefficient and local efficiency were associated with poorer auditory comprehension, memory, and reasoning. Lesion volume was associated with worse language and cognitive outcomes but was not associated with global cerebellar network measures after covariate adjustment. These findings provide novel network-level evidence of altered cerebellar topology in PSA, extending prior cortical-focused models. The results suggest that cerebellar network disruption is associated with language and cognitive deficits and may inform the development of network-based imaging markers and motivate future longitudinal and intervention studies in aphasia.