Effective Connectivity Of The Motor-related Cortex In Stroke:a Resting-state FMRI Study With Granger Causality Analysis

Backgrounds and Purpose:Hand motor dysfunction has a particularly serious negative impact on daily activities and quality of life in stroke patients. Resting-state fMRI is popular for revealing the neural mechanism of stroke patients with motor dysfunction, since its performing process is simple and patients are easy to cooperate. The abnormal resting-state functional connectivity of the primary motor cortex (M1) which is the key brain region controlling hand movement, is generally considered to reflect the cerebral cortex function reorganization of stroke patients with hand motor dysfunction. However, the common functional connectivity analytical methods of resting-state functional Magnetic Resonance Imaging (fMRI) are unable to reveal the direction of functional connectivity in the involved brain regions. The method of effective connectivity analysis can reflect the directional connectivity between different brain regions or nervous systems. So, the purpose of this study is to observe whether effective connectivity of the ipsilesional Ml is disturbed in stroke patients, who show different outcomes in hand motor function, and further to investigate whether the abnormal effective connectivity is related to the severity of hand dysfunction.Methods:23 patients with unilateral left hemisphere subcortical stroke and 24 matched healthy controls (HC) were selected and performed resting-state fMRI scans. Meanwhile, the stroke patients were divided into two subgroups:partially paralyzed hands (PPH) and completely paralyzed hands (CPH) according to their practical abilities in hand function. A voxel-wise Granger causality analysis (GCA) on resting-state fMRI data between the ipsilesional M1 and the whole brain was performed to explore differences between the PPH, the CPH and the HC.Results:The effective connectivity from the ipsilesional Ml to the sensorimotor cortices decreased greater in CPH group than that in PPH group. The influence from the frontoparietal cortices to the ipsilesional M1 diminished in both stroke subgroups. Compared with the HC group, the PPH group showed significantly decreased effective connectivity from the ipsilesional M1 to the ipsilesional cerebellum; while the CPH group exhibited decreases in effective connectivity from the ipsilesional M1 to bilateral cerebellum and the contralesional ventral premotor cortex, in contrast, increases in effective connectivity from the ipsilesional M1 to the ipsilesional superior frontal gyrus and occipital lobe. In the opposite direction, compared with the HC group, the PPH group showed a decrease in effective connectivity from the bilateral anterior cingulate gyrus (ACC) to the ipsilesional M1,but the CPH group only showed increased effective connectivity from the contralesional cerebellum to the ipsilesional M1. Compared with the PPH group, the CPH group displayed decreased effective connectivity from the ipsilesional Ml to the contralesional cerebellum and from the contralesional superior parietal lobe (SPL) to the ipsilesional M1, and the strength of the connectivities were positively correlated with the FMA scores; conversely, the CPH group showed increased connectivities from the ipsilesional M1 to the ipsilesional middle temporal gyrus and middle frontal gyrus, and the intensity of connectivities were negatively correlated with the FMA scores.Conclusions:Our results indicated that the stroke patients had an abnormal effective connectivity network of the ipsilesional M1, which was related to the severity of hand dysfunction. These findings provided more valuable information for improving our understanding of the role of brain areas related to motor-execution and frontoparietal motor control networks during brain plasticity following stroke, and provided reliable evidence and guidance for using appropriate clinical rehabilitation.