Functional connectivity between non-motor and motor networks predicts motor recovery changes after stroke.
Authors
Affiliations (7)
Affiliations (7)
- Department of Neurosurgery, The First Hospital of Jilin University, Changchun, Jilin, China.
- State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, 100875, China.
- Beijing Key Laboratory of Brain Imaging and Connectomics, Beijing Normal University, Beijing, China.
- Chinese Institute for Brain Research, Beijing, China.
- Department of Neurology, The First Affiliated Hospital, Sun Yat-Sen University, Guangdong Provincial Key Laboratory for Diagnosis and Treatment of Major Neurological Diseases, National Key Clinical Department and Key Discipline of Neurology, Guangzhou, China. [email protected].
- Department of Neurosurgery, The First Hospital of Jilin University, Changchun, Jilin, China. [email protected].
- Department of Neurology, The First Affiliated Hospital, Sun Yat-Sen University, Guangdong Provincial Key Laboratory for Diagnosis and Treatment of Major Neurological Diseases, National Key Clinical Department and Key Discipline of Neurology, Guangzhou, China. [email protected].
Abstract
Stroke impairs limb motor function, which affects patients' quality of life and imposes economic burdens. Early prediction of motor recovery is essential for guiding treatment and rehabilitation. While the corticospinal tract is a known biomarker, the role of non-motor brain regions remains under explored. Fifty-five stroke patients with unilateral subcortical lesions and 49 healthy controls underwent resting-state functional MRI scans at 1 week, 4 weeks, and 12 weeks after stroke. Focusing on two motor and 15 non-motor networks defined by the Schaefer atlas, machine learning models were used to predict changes in motor function measured by the Fugl-Meyer assessment using functional connectivity (FC) data. The network-based statistic (NBS) method was used to identify significant FC differences between patients and controls. Among 90 predictive models tested, only the model based on FC within the Somatomotor A (SomMotA) and Control A (ContA) networks at 1 week after stroke significantly predicted motor recovery from the acute to subacute phases (p = 0.00040 after Bonferroni correction). The ContA network contributed more to the prediction than the SomMotA network did. NBS analysis revealed significant FC alterations within the SomMotA network in patients versus controls but no direct correlation between predictive FC and group differences. This study revealed acute-phase FC between the non-motor ContA and motor SomMotA networks can be used to effectively predict motor recovery in stroke patients. These findings highlight the significant role of non-motor networks in motor recovery and suggest that rehabilitation strategies incorporating non-motor interventions may improve patient outcomes.