Dynamic functional network connectivity remodeling in cervical spondylotic myelopathy: insights into postoperative neural recovery.

Abstract

The longitudinal changes in large-scale brain network dynamic functional network connectivity (dFNC) and their role in postoperative recovery remain insufficiently explored. To investigate the remodeling of brain dFNC in individuals with cervical spondylotic myelopathy (CSM), focusing on temporal characteristics and their association with neural function recovery. Cross-sectional study. The study included 32 CSM patients and 32 age- and sex-matched healthy controls (HCs). We calculated the dFNC states and their temporal characteristics, and the correlation of these measures with improvements in clinical symptoms were assessed as key outcomes. Group independent component analysis (GICA) was employed to extract whole-brain independent components (ICs). A sliding time window and k-means clustering were utilized to identify dFNC states. Intergroup differences in connectivity were systematically compared, and correlation analyses were conducted to associate temporal variations in dFNC with clinical recovery outcomes. GICA identified ten functional networks, and dFNC revealed four distinct states. Participants predominantly occupied State 1, indicated by higher mean dwell time and fractional time. Preoperatively, CSM patients showed reduced functional connectivity (FC) in the visual, default mode, and frontoparietal networks. Three months postoperatively, these patients partially regained functional connectivity in some dynamic states. Additionally, changes in fractional time (FT) in State 4 were significantly negatively correlated with improvements in neural function. This study offers a dynamic perspective on the remodeling of large-scale brain networks in patients with CSM following surgery. These findings elucidate the neurobiological mechanisms underlying spinal cord recovery postdecompression and suggest novel therapeutic strategies for postoperative rehabilitation.