Objective This study aimed to explore the differences in global and local brain network topological properties between deficient pattern(DP)and excess pattern(EP)of mild vascular cognitive impairment caused by subcortical small vessel disease based on graph theory network.Methods Patients were recruited prospectively and were classified with DP and EP subtype.The global small-world topological attributes and local nodes were calculated for the comparison of DP,EP,and healthy controls(CN)using the GRETNA platform.Results The three groups all had small-world attributes,but only the patients in EP had a significantly lower small world attribute δ in the range of 0.05-0.26 than the control group(P<0.05).The node efficiency and node strength indicators of multiple brain region were able to significantly distinguish the DP group from the EP group.However,there was no positive brain region in the node efficiency of the DP patients(P>0.05),and only a few brain regions showed increased node strength efficiency(P<0.05).Conclusion The results indicate that the syndrome of DP and EP have significantly different neuroimaging phenotypes,providing a basis for further research of biological classification based on Chinese Medicine syndromes.