This paper uses resting-state functional magnetic resonance imaging (rs-FMRI) to construct a whole-brain binary functional network through a complex brain network analysis theory based on graph theory to explore the functional network of patients with type 2 diabetes (T2DM). Changes in topological properties and their potential relationships with fasting blood glucose (FBG), glycated haemoglobin (HbAlc), and cognitive function scale, and further explore the diagnostic value of rs-FMRI technology for central nervous system damage in T2DM patients, for clinical diagnosis and treatment Provide objective radiological evidence. In the range of sparsity (Sp) of 0.05 to 0.50 and a step size of 0.01, compared with the random network, the resting brain functional networks in the T2DM group and the HC group have larger clustering coefficients and similar shortest paths. Length and small world index greater than 1, that is, both groups of resting brain functional networks have small world characteristics. The MoCA score of the T2DM group was positively correlated with the node degree (r = 0.400, p = 0.043) and the node efficiency (r = 0.452, p = 0.021) of the right straight back. FBG is positively correlated with the node degree of the left occipital gyrus (r = 0.422, p = 0.023); HbAlc is related to the node degree of the left occipital gyrus (r = 0.372, p = 0.043) and the node degree of the left occipital gyrus ( r = 0.382, p = 0.037) was positively correlated with the node intermediary (r = 0.388, p = 0.034) at the back of the right cingulate gyrus. The topological properties of the resting brain function network of T2DM patients with negative MRI findings have changed compared with normal people, indicating that T2DM is an important factor leading to brain function damage, further explaining the rs-fMRI technology and complex brain networks based on graph theory Analysis theory can be used as an effective method to study the changes of brain function in T2DM patients.

本文采用静息态功能磁共振成像（rs-FMRI），通过基于图论的复杂脑网络分析理论构建全脑二元功能网络，以探索2型糖尿病（T2DM）患者的功能网络。拓扑特性的变化及其与空腹血糖（FBG），糖化血红蛋白（HbAlc）和认知功能量表的潜在关系，并进一步探索rs-FMRI技术对T2DM患者中枢神经系统损伤的诊断价值，用于临床诊断和治疗提供客观的放射学证据。与随机网络相比，在稀疏度（Sp）为0.05至0.50且步长为0.01的范围内，T2DM组和HC组的静息大脑功能网络具有较大的聚类系数和相似的最短路径。长度和小世界指数大于1，即两组静止的大脑功能网络都具有小世界特征。T2DM组的MoCA评分与右直背的节点度（r = 0.400，p = 0.043）和节点效率（r = 0.452，p = 0.021）正相关。FBG与左枕回的结节度呈正相关（r = 0.422，p = 0.023）；HbAlc与左枕回的结节度有关（r = 0.372，p = 0.043），左枕回的结节度（r = 0.382，p = 0.037）与结节中介度呈正相关（r = 0.388） ，p = 0.034）在右扣带回的后面。