【作者】 吴晶涛；

【导师】 杨海山；

【作者基本信息】 吉林大学 ， 放射医学， 2014， 博士

【摘要】 背景功能磁共振（fMRI）利用检测脑血氧代谢水平的变化间接对脑功能活动成像，如认知功能、注意功能、执行功能、感觉功能等；具有同步低频fMRI信号的空间远隔脑区组成脑网络，意味着这些脑区能够协同活动。人脑网络是一个非常复杂的网络构造；脑默认网络是被最早发现的脑网络之一，大量的研究表明网络受年龄、衰老、神经精神疾病影响，生理性脑衰老过程蕴含网络功能连接性减低。目的静息状态脑功能磁共振网络包含了时空信号相反的网络构造，即认知任务活动诱导的正性激活的网络和负激活的默认网络。默认网络、背侧注意网络、执行控制网络和突显网络构成的大尺度网络是支撑认知活动的基础；本课题进一步研究默认网络子网络的正负相关网络及默认网络、背侧注意网络、执行控制网络和突显网络构成的大尺度网络在老年期的退变，特别关注这些网络变化的程度。方法对18位正常青年人和22位健康老年人进行了静息功能磁共振扫描，成像序列包括常规结构成像及血氧水平依赖（BOLD）的T2*加权像。检查前对所有受试者进行神经心理学测评，其简易精神状态量表评分28-30分。图像处理应用国际通用脑功能图像处理软件，以后扣带回（PCC）和前额叶内侧腹侧区（vMPFC）为种子区，分析功能磁共振图像低频波动信号，基于全脑像素提取默认网络成分及与其呈反相关关系的网络，对两组受试者网络连接性进行组内和组间统计学分析。另外，利用种子区信号相关性分析提取功能磁共振图像中的四个制式网络，包括默认网络（DMN）、背侧注意网络（DAN）、执行网络（ECN），突显网络（SN），以感觉系统视觉网络作为对照控制，比较大尺度网络在老年组与青年组的变化。结果组内统计分析采用单样本t检验（双侧），取阈值p＜0.01，经Monte-Carlo模拟校正，激活聚丛大小18像素（486mm3），校正后像素p＜0.01，分别获得了青年组和老年组的PCC+、vMPFC+、PCC-和vMPFC-网络。组间比较分析采用双样本t检验（双侧），取阈值p＜0.05，经Monte-Carlo模拟校正，激活聚丛大小54像素（1458mm3），校正后像素p＜0.05。老年组的默认网络及其反相关网络表现了以功能连接性减低为特征的变化。构成默认网络的主要脑区呈现了与PCC及vMPFC的功能连接性减低；PCC负相关和vMPFC负相关网络功能连接性减低的脑区分布具有不同特点，前者位于两侧额顶叶侧面，后者以大脑内侧辅助运动区为主。老年脑功能网络的损害分布呈前后方向梯度改变，前部较后部明显，与以往研究结果一致。两组数据样本t检验，在经alphasim校正的p<0.01的阈值水平下，显示DMN受影响最显著，其次为DAN，再次为ECN，SN受影响最小；默认网络、背侧注意网络、执行控制网络和突显网络分别损失了14.9%，7.9%，4.6%及1.6%的功能连接区。视觉网络作为控制因素，经统计比较我们发现老年组和青年组的视觉网络无明显差异。同时，利用基于体素的形态测量（VBM）比较了青年组和老年受试者的脑灰质体积，结果显示老年人的额叶、顶叶及颞叶的内侧面、外侧面灰质体积呈现广泛的显著减少，而枕顶叶体积无显著变化。结论老年人注意网络和默认网络表现了功能连接性减低为特征的变化，这些网络改变有助于更深入理解老年期脑功能生理性衰退；正常人的脑衰老表现为在网络水平上呈现有组织性的变化；我们的研究表明老年期大尺度的脑网络呈现选择性的损害，高级认知网络较低级脑功能网络退变更突出。静息功能磁共振网络分析是无创性探索老年脑内在机制变化的重要工具，脑网络的变化可以作为研究老年相关脑疾病的潜在生物学标志。更多还原

【Abstract】 Background Functional MRI indirectly reflect brain activitiessupporting cognition, sensory, motor and so on, by the way of bloodoxygen level depend (BOLD). Brain networks organized by spatialdistance brain regions with fMRI signals of synchronic low frequencyfluctuations mean they could function cooperatively. Age-relatedhigh-order brain functions such as memory, attention, problem solving aswell as sensorimotor ability may alter at different degree, the intrinsicbrain networks underlying the brain functions remain largely unclear,aging-related changes in aging brain should be investigated further.Objective Intrinsic brain activity in a resting state incorporatescomponents of the task negative network called default mode network(DMN) and task-positive networks called attentional networks.Attentional function impairment is one of the major clinicalmanifestations in elderly people. Anti-correlation is a critical feature ofbrain networks. In the present study, the subnetworks of DMN and theircorresponding anti-correlated networks were compared between the eldergroup and young group to investigate the differences of the intrinsic brainactivity. Another aim of our study was to investigate the patterns of alterations in the large-scale brain networks with aging progress,including default mode network (DMN), executive-control network(ECN), dorsalattentional network (DAN) and salience network; we areespecially concerned of the degree to which the large-scale intrinsicnetworks disruted respectively.Methods Two groups of healthy subjects including18young and22older adults were performed resting-state functional magnetic resonanceimaging (fMRI) scanning with a blood oxygen level dependent sensitivesequence (BOLD). All the subjects were normal in the neuropsychologytest with MMSE scores ranged from28to30. Four network componentsof positive and negative correlations were extracted based on seedregions of posterior cingulate cortex (PCC) and ventromedial prefrontalcortex (vMPFC) from the fMRI data, respectively. The fMRI data wereprocessed by international widely used software. Four canonicalresting-state networks including the DMN, ECN, DAN and saliencenetwork were determined by employing seed-based temporal analysis.The visual network was used as control. Connectivity maps werecompared between groups by using two-sample t tests with a thresholdadjustment method based on Monte-Carlo simulations correction at avoxelwise p<0.05.Results Characteristics of decreased functional connectivity in both ofpositive and negative subetworks of the DMN were found in the older group. The traditional regions involved in the DMN showed significantconnectivity decrease. Changes of PCC negative correlations map andvMPFC negative correlations map presented different patterns indistribution. The vMPFC negative network displayed attenuated activitymainly in midline regions (mainly containing supplementary motor area,SMA), whereas the PCC negative network demonstrated decreasedactivity in the regions of lateral fronto-parietal areas.We found that distinct disruptive alterations presented in thelarge-scale aging-related resting brain networks. That is, the DMN wasimpaired the mostly, followed by the DAN and ECN, the saliencenetwork showed minimally functional connectivity disruption. For theelderly group, network volumes reduced by14.9%,7.9%,4.6%and1.6%when compared with each of unified mask for DMN, DAN, ECN and SN,respectively. The visual network was used as control measurement of thehigh order networks, and was found equivalently preserved as control.Voxel based morphology (VBM) analysis indicated extensive braingray matter atrophy in lateral and medial frontal lobe, temporal lobe andparts of parietal lobe, whereas the volume of the parieto-occipital loberelatively preserved in the aged subjects.Conclusions The decreased functional connectivity in brain networks inthe old group may reflect impairments of cognitive functions evenwithout any task load or stimuli. Degeneration of brain in normal people altered organizationally a level of networks. Our findings suggested theaged brain is characterized of selective vulnerability in large-scale brainnetworks. The findings could help for understanding the degenerationmechanism in aging brain. Resting-state fMRI could be an importantnon-invasive tool for investigating the mechanisms in aging brain.更多还原