目的:探讨脑白质病变(white matter lesion，WML)合并不同程度认知障碍患者与正常人默认网络(default mode network，DMN)、额顶叶网络(frontoparietal network，FPN)、凸显网络(salience network，SN)、注意网络(dorsal attention network，DAN)之间的功能连接(functional connectivity，FC)改变，以及这种功能连接变化与执行功能评分的相关性。方法:选择2016年1月至2020年12月在北京天坛医院就诊的WML患者，分为WML-非痴呆血管性认知障碍组(none-dementia vascular cognition impairment，VCIND组)(n n＝19)，WML-血管性痴呆组(vascular dementia，VaD组)(n n＝17)，同时选取无认知功能障碍且年龄、性别、受教育程度相匹配的正常对照组(normal control，NC组)(n n＝24)。使用Stroop色词干扰测验B、Stroop色词干扰测验C、连线测验B(trail making test B，TMT-B)、言语流畅性测验(verbal fluency test，VFT)评估所有受试者的执行功能。使用独立成分分析(independent component analysis，ICA)方法挑选出FPN、SN、DMN、DAN四个静息态网络。采用SPSS 20.0软件进行统计分析，使用协方差分析比较三组间这四个静息态网络所包含重要脑区间功能连接差异，使用Spearman相关分析对这种功能连接差异与执行功能评分做相关性分析。n 结果:(1)VaD组内侧前额叶皮质(medial prefrontal cortex，MPFC)-左侧顶下小叶(left inferior parietal lobe，IPL.L)之间的FC(0.267±0.320)较NC组(0.520±0.253)显著降低(n P＝0.005)；VaD组MPFC-后扣带回(posterior cingulate，PCC)的FC(0.655±0.284)较VCIND组(0.810±0.232)显著下降(n P＝0.017)；VaD组MPFC-左侧岛叶(left insula，insula.L)、MPFC-左侧背外侧前额叶皮质(left dorsolateral prefrontal cortex，DLPFC.L)之间的FC[(0.411±0.277)，(0.545±0.311)]较VCIND组[(0.239±0.308)，(0.353±0.270)]显著升高(n P＝0.044，0.028)；VCIND组左侧顶上小叶(left superior parietal lobe，SPL.L)-右侧背外侧前额叶皮质(right dorsolateral prefrontal cortex，DLPFC.R)之间的FC(0.488±0.157)较NC组(0.301±0.257)显著升高(n P＝0.010)；VaD组MPFC-insula.L、MPFC-DLPFC.L之间的FC较NC组显著升高(n P＝0.020，0.037)。(2)MPFC-PCC之间的功能连接差异与Stroop色词干扰C(n r＝-0.279，n P＝0.036)、TMT-B(n r＝-0.313，n P＝0.018)评分存在负相关；MPFC-IPL.L之间的功能连接差异与Sroop色词干扰B(n r＝-0.311，n P＝0.018)、Stroop色词干扰C(n r＝-0.308，n P＝0.020)及TMT-B(n r＝-0.367，n P＝0.005)评分存在负相关，与VFT存在正相关(n r＝0.357，n P＝0.006)；SPL.L-DLPFC.R之间的功能连接差异与Stroop色词干扰B评分呈正相关(n r＝0.305，n P＝0.021)。n 结论:WML患者认知相关脑网络内部脑区之间的功能连接存在一定的差异，不同脑区间功能连接的差异与执行功能存在相关性，这些脑区间功能连接的变化可能在一定程度上预示着可能出现的执行功能的下降，这在一定程度上可解释患者出现认知功能障碍的机制。“,”Objective:To investigate the cognitive-related networks of patients with different cognitive impairment related to white matter lesions (WMLs), and explore the correlation between the altered functional connectivity and the executive function.Methods:Patients with white matter lesions in Beijing Tiantan Hospital from January 2016 to December 2020 were selected , who were divided into WML-non-dementia vascular cognitive impairment (WML-VCIND) group(n n=19)and WML-vascular dementia (VaD) group(n n=19). At the same time, 24 normal control group (NC) with no cognitive impairment with age, gender and education level matched were enrolled. The Stroop B, Stroop C, trail making test(TMT-B) and verbal fluency test (VFT) were used to evaluate the executive function of all subjects. The ICA method was used to identity four cognitive-related RSNs: frontoparietal network (FPN), salience network (SN), default network (DMN), and dorsal attention network (DAN). Using SPSS 20.0 software, the functional connectivity (FC) of the regions of interest of the selected RSNs among the three groups were compared by analysis of covariance, and correlation between the altered FCs and executive function scores were explored by Spearman correlation analysis.n Results:(1) The FC of the medial prefrontal cortex (MPFC)-left inferior parietal lobe (IPL.L) in the VaD group(0.267±0.320) was significantly lower than that in the NC group (0.520±0.253)(n P=0.005). The FC of MPFC-posterior cingulate (PCC) was significantly lower in the VaD group(0.655±0.284) than that in the VCIND group(0.810±0.232) (n P=0.017). The FC of MPFC-left insula and MPFC-left dorsolateral prefrontal cortex (DLPFC.L)was significantly higher in the VaD group(0.411±0.277, 0.545±0.311)than that in the VCIND group(0.239±0.308, 0.353±0.270)(n P=0.044, 0.028). The FC of the left superior parietal lobe (SPL.L)-right dorsolateral prefrontal cortex (DLPFC.R) in the VCIND group(0.488±0.157) was significantly higher than that in the NC group(0.301±0.257) (n P=0.010). The FC of MPFC-left insula and MPFC-DLPFC.L was higher in NC group than that in VaD group (n P=0.020, 0.037). (2) The FC of MPFC-PCC was negatively correlated with Stroop C score (n r=-0.279, n P=0.036), and TMT-B score (n r=-0.313, n P=0.018). The FC of MPFC-IPL.L was negatively correlated with Stroop B score (n r=-0.311, n P=0.018), Stroop C score(n r=-0.308, n P=0.020) and TMT-B score (n r=-0.367, n P=0.005), while positively correlated with VFT score (n r=0.357, n P=0.006). The FC of SPL.L-DLPFC.R was positively correlated with the Stroop B score (n r=0.305, n P=0.021).n Conclusion:There are certain differences in the functional connectivity between brain areas within the cognitive related brain networks, and the differences in FC between different brain areas are correlated with executive function scores. The changes of the FC may indicate possibly decline in executive function, which can explain the mechanism of cognitive declines.