Objective:To explore the changes in topological attributes of structural covariance network based on cortical thickness and the brain functional activities in patients with anti-N-methyl-D-aspartate receptor (NMDAR) encephalitis by graph theory and functional connectivity (FC) analyses, and to investigate whether these changes were correlated to cognitive impairment.Methods:A total of 33 patients with anti-NMDAR encephalitis from Department of Neurology of the First Affiliated Hospital of Guangxi Medical University(patient group) and 35 healthy controls(control group) with matched gender, age, and education were included from July 2018 to November 2021.All subjects received cognitive function assessments, structural and functional magnetic resonance imaging scans.Structural covariance networks were constructed in the two groups based on cortical thickness values and topological characteristics of networks were computed.A non-parametric permutation test which repeated 1 000 times was used to compare the characteristics of the networks between the two groups.Brain regions with abnormal topology were defined as region of interest(ROI), and FC values in global brain level were calculated.SPM 12 and RESTplus were used to identify the brain regions with significant differences in FC values between the two groups.Finally, Spearman correlation analysis between FC values of significant brain regions and cognitive scores were performed by SPSS 24.0.Results:The cognitive score of patients with anti-NMDAR encephalitis (27.0(23.5, 28.0)) was lower than that in control group(29.0(27.0, 30.0)) ( Z=-3.029, P=0.002). Graph theory analysis found that the patients showed significantly increased clustering coefficients ( P=0.004) and decreased global efficiency ( P=0.004) compared with healthy controls.Moreover, the nodal efficiency of left ventral posterior cingulate cortex (vPCC) and right dorsal posterior cingulate cortex (dPCC), as well as the nodal degree centrality of left vPCC and left polar planum of superior temporal gyrus (ppSTG) in patient group were significantly decreased ( P<0.05, FDR corrected) compared with control group.FC analysis showed the increased FC values between left vPCC and posterior cerebellum (MNI: x=6, y=-66, z=-21), as well as between left ppSTG and anterior cerebellum (MNI: x=6, y=-54, z=-12) (GRF corrected, voxel level P<0.001, cluster level P<0.05) in patient grooup.The FC values between left vPCC and posterior cerebellum were negatively correlated with the cognitive scores ( r=-0.403, P=0.020). Conclusion:Patients with anti-NMDAR encephalitis show abnormal topology of structural covariance network based on cortical thickness and altered FC values, some of which are correlated to cognition and may be the underlying neural mechanism of cognitive impairment in patients with anti-NMDAR encephalitis.