A Study on the Topological Properties of Brain Networks in Patients with Hemiplegic Shoulder Pain Based on Functional Magnetic Resonance Imaging

Background : Hemiplegic shoulder pain (HSP), a common complication of stroke, has a high incidence (34%-85%) and a complex mechanism. The aim of this study is to explore the differences of brain network topological properties between patients with and without shoulder pain after stroke with hemiplegia using resting-state functional magnetic resonance imaging. Method : Fifty patients with hemiplegia after stroke were divided into a shoulder pain group (SP) and non pain group (NP) according to the presence or absence of shoulder pain symptoms, and they underwent resting-state functional magnetic resonance imaging. GRETNA was used to calculate the global and regional network topological features to compare the differences in brain network between the two groups Results : In terms of the global brain network, the gray matter network of hemiplegic patients had obvious "small-world" property (σ> 1), Global efficiency was decreased in SP (0.1733±0.0047) compared with NP (0.1765±0.0044)(P=0.02), the characteristic path length was increased in SP (0.3224±0.0152) compared with NP (0.3170±0.0089)(P=0.001). In terms of the regional brain network, in SP there was reduced in Node degree centrality and Node local efficiency in the right frontal inferior operculum gyrus and orbital gyrus compared with the NP (P<0.05). The SP showed a decrease in the Nodal betweenness centrality of the right pallidum, left dorsolateral superior frontal gyrus, left frontal inferior triangular gyrus, and left rolandic opercular compared to the NP (P<0.05). The SP exhibited increased Nodal betweenness centrality in the left thalamus, right parahippocampal gyrus, left inferior occipital gyrus, and left angular gyrus compared to the NP (P<0.05) Conclusion : The topological properties of brain networks in hemiplegic shoulder pain patients have shifted to a "weaker small world" pattern; The changes of nodes are mainly concentrated in the executive control network, default mode network, basal ganglia region and language network, where related to pain and uncontrolled emotion regulation, provides new insights for the related research of hemiplegic shoulder pain after stroke.