Longitudinal Changes in Brain Network Metrics and Their Correlations with Spinal Cord Diffusion Tensor Imaging Parameters Following Spinal Cord Injury and Regenerative Therapy

<b>Objectives:</b> Spinal cord injury (SCI) disrupts the microstructure of the spinal cord, triggers reorganization of the brain network, and causes motor deficits. However, the temporal dynamics and interrelationships of these alterations remain unclear. <b>Methods:</b> Eight monkeys underwent spinal cord hemisection and were randomly assigned to either the SCI-only group or the treatment group that received neurotrophin-3-chitosan implants. Longitudinal brain structural/resting-state magnetic resonance imaging and spinal cord diffusion tensor imaging (DTI) were conducted. Concurrently, hindlimb motor function was assessed. The brain network topology was characterized through graph theory. The generalized additive mixed model (GAMM) was employed to analyze the longitudinal trajectories of network metrics, while the linear mixed-effects model (LMM) was used to evaluate the moderating effect of treatment on correlations between network metrics and motor/DTI parameters. <b>Results:</b> The SCI-only group exhibited sustained functional network segregation, aberrant structural topology, and lower fractional anisotropy (FA). These findings collectively reflect chronic maladaptive plasticity. In the treatment group, the therapy not only enhanced white matter integrity, reflected by increased FA values, but also reduced the clustering coefficient (Cp) in brain structural network, indicating a shift away from maladaptive segregation. Critically, the LMMs further revealed that treatment moderated the pathological correlations between global efficiency (Eg), local efficiency, Cp, and locomotor parameters. Moreover, spinal FA exerted a significant main effect on Eg of brain functional networks. <b>Conclusions:</b> These findings suggest that treatment-induced brain reorganization underlies motor function following SCI, and progressive brain reorganization correlates with changes in spinal cord microstructure, revealing a systems-level mechanism of neural repair.