Structural and EEG Motor Networks Distinguish Level of Motor Impairment After Stroke
European Journal of Neuroscience · 2026-11 · 64 (9)
DOI: 10.1111/ejn.70747 ↗Abstract
ABSTRACT The objective of this study was to determine how structural and EEG‐based functional connectivity jointly distinguish levels of motor impairment after stroke. We estimated patient‐specific (1) structural connectivity using MRI‐derived lesion models and (2) functional connectivity using source‐localized EEG, guided by each patient's structural connectome. Analyses focused on the brainstem and three motor‐related cortical regions bilaterally: primary motor cortex (M1), dorsal premotor cortex (PMd), and supplementary motor cortex (SMA). Network analysis metrics (degree centrality, average efficiency, betweenness centrality) were applied to structural and functional data, contrasting motor networks between patients with high (H) and low (L) upper extremity motor function. Structural disconnection of the brainstem from M1 and PMd carried the highest probability of low motor status. All network measures were elevated in the H group compared to the L group across motor‐related regions, indicating a globally strengthened motor network organization associated with better motor function. Functional connectivity revealed convergent but distinct patterns. In the delta band (1–3 Hz), ipsilesional motor regions showed stronger connectivity in the H group. In the beta band (14–29 Hz), contralesional motor regions demonstrated enhanced connectivity in H patients, consistent with prior evidence that contralesional motor regions may contribute to motor function after ipsilesional injury. We also introduced a novel metric, motor betweenness centrality, to quantify the role of nonmotor regions in rerouting motor‐related communication. Almost all new hubs identified structurally were associated with lower motor status, suggesting less efficient network organization in more impaired patients. Thus, preservation of direct motor pathways supports better motor status, while reliance on alternative, less efficient routes correlates with worse outcomes. By contrast, EEG routing revealed that H patients recruited proximal cortical hubs, while L patients relied on distant, inefficient ones, highlighting functional network organization not evident in the structural analysis. These patterns suggest that more local network organization, rather than more distributed rerouting, is associated with better motor status within the motor system. In summary, multimodal network properties reliably identified key motor ROIs predictive of motor status. This concordance suggests that EEG‐based metrics, particularly delta‐ and beta‐band connectivity, are associated with lesion‐derived structural disruption and complementary features of motor‐network organization. Structural connectivity damage within motor circuits defines the extent of network disruption associated with motor impairment, while EEG functional connectivity provides complementary information about network organization measured at the time of assessment. Together, these measures characterize different aspects of motor network integrity after stroke.
EUVIMED