AUTHOR=Hou Yanbing , Zhang Lingyu , Wei Qianqian , Ou Ruwei , Yang Jing , Gong Qiyong , Shang Huifang 

TITLE=Impaired Topographic Organization in Patients With Idiopathic Blepharospasm

JOURNAL=Frontiers in Neurology

VOLUME=12

YEAR=2022

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2021.708634

DOI=10.3389/fneur.2021.708634

ISSN=1664-2295

ABSTRACT=<p><bold>Background:</bold> Idiopathic blepharospasm (BSP) is a common adult-onset focal dystonia. Neuroimaging technology can be used to visualize functional and microstructural changes of the whole brain.</p><p><bold>Method:</bold> We used resting-state functional MRI (rs-fMRI) and graph theoretical analysis to explore the functional connectome in patients with BSP. Altogether 20 patients with BSP and 20 age- and gender-matched healthy controls (HCs) were included in the study. Measures of network topology were calculated, such as small-world parameters (clustering coefficient [<italic>C</italic><sub>p</sub>], the shortest path length [<italic>L</italic><sub>p</sub>]), network efficiency parameters (global efficiency [<italic>E</italic><sub>glob</sub>], local efficiency [<italic>E</italic><sub>loc</sub>]), and the nodal parameter (nodal efficiency [<italic>E</italic><sub>nod</sub>]). In addition, the least absolute shrinkage and selection operator (LASSO) regression was adopted to determine the most critical imaging features, and the classification model using critical imaging features was constructed.</p><p><bold>Results:</bold> Compared with HCs, the BSP group showed significantly decreased <italic>E</italic><sub>loc</sub>. Imaging features of nodal centrality (<italic>E</italic><sub>nod</sub>) were entered into the LASSO method, and the classification model was constructed with nine imaging nodes. The area under the curve (AUC) was 0.995 (95% CI: 0.973–1.000), and the sensitivity and specificity were 95% and 100%, respectively. Specifically, four imaging nodes within the sensorimotor network (SMN), cerebellum, and default mode network (DMN) held the prominent information. Compared with HCs, the BSP group showed significantly increased <italic>E</italic><sub>nod</sub> in the postcentral region within the SMN, decreased <italic>E</italic><sub>nod</sub> in the precentral region within the SMN, increased <italic>E</italic><sub>nod</sub> in the medial cerebellum, and increased <italic>E</italic><sub>nod</sub> in the precuneus within the DMN.</p><p><bold>Conclusion:</bold> The network model in BSP showed reduced local connectivity. Baseline connectomic measures derived from rs-fMRI data may be capable of identifying patients with BSP, and regions from the SMN, cerebellum, and DMN may provide key insights into the underlying pathophysiology of BSP.</p>