基于动态因果模型的轻度认知功能障碍工作记忆脑电效应连接模式研究

发布时间：2018-06-03 10:02

本文选题：轻度认知功能障碍 + 工作记忆　；参考：《天津医科大学》2017年硕士论文

【摘要】：研究目的:轻度认知功能障碍(Mild Cognitive Impairment,MCI)是介于正常衰老和阿尔兹海默症(Alzheimer’s Disease,AD)之间的一种认知障碍症候群,工作记忆障碍是MCI的主要临床表现。研究MCI的工作记忆障碍机制具有重要的意义。目前,脑网络研究是神经科学领域研究的热点。本论文从脑网络的角度,对MCI患者在工作记忆任务下的多通道脑电(multi-channel electroencephalographs,EEGs)数据进行动态因果模型(Dynamic Causal Modeling,DCM)分析,研究MCI患者工作记忆脑效应网络的连接模式,为探索MCI的工作记忆障碍机制提供新的视角。研究方法:1.受试者和实验数据:MCI组、正常老年组(Elderly组)和正常青年组(Youth组)各15名受试者,分别记录每位受试者在视觉工作记忆任务下的行为学数据和34通道脑电数据。2.行为学分析:计算MCI组、Elderly组和Youth组在工作记忆任务下的平均反应时间和正确率。3.数据预处理:对采集到的EEGs去除工频干扰、基线漂移、眼动干扰和肌电干扰。4.时频分析:应用短时傅里叶变换(STFT)方法计算预处理后的EEGs的能量密度分布,分析工作记忆EEGs的特征频段和特征时段。5.源分析:应用分布式线性模型的方法确定MCI组、Elderly组和Youth组在工作记忆特征时段和频段上的激活脑区。6.DCM分析:根据源分析所确定的工作记忆激活脑区,构建DCM网络,计算三组受试者DCM网络子连接的超越概率(Exceedance Probability,EP),分析工作记忆脑效应网络的连接特性;分别计算三组受试者DCM网络中不同方向连接族(正向连接、反向连接和横向连接)和左右侧连接族的超越概率,比较分析MCI组工作记忆信息的传递方向和大脑偏侧性。7.统计学分析:在SPSS 22.0统计分析软件上应用单因素方差分析方法对实验数据进行统计学检验。研究结果:1.行为学结果MCI组、Elderly组和Youth组的工作记忆任务平均反应时间分别为:1387.8±80.0ms、1262.1±69.9ms、912.7±66.4ms,MCI组的反应时间显著大于Elderly组(P0.01)和Youth组(P0.001);MCI组、Elderly组和Youth组的正确率分别为:73.38±1.10%、82.58±1.42%、92.53±1.49%,MCI组的正确率显著低于Elderly组(P0.01)和Youth组(P0.001)。2.工作记忆EEGs的能量密度三组受试者工作记忆EEGs的能量密度主要集中在theta频段(4-8Hz),MCI组EEGs在theta频段上的能量密度显著低于Elderly组和Youth组。三组受试者的工作记忆特征通道的时频图表明工作记忆延迟期的能量密度最为连续和集中。3.工作记忆激活脑区源分析结果显示,MCI组的激活脑区为双侧额中线皮层(middleline-frontal cortex,MFC);Elderly组的激活脑区为双侧背外侧前额叶皮层(dorsolateral prefrontal cortex,DLPFC)和双侧MFC;Youth组的激活脑区为双侧DLPFC、双侧MFC和双侧顶叶皮层(parietal cortex,PAR)。将Youth组的激活脑区作为网络节点进行DCM分析。4.DCM分析结果(1)DCM网络子连接的超越概率MCI组:双侧PAR到同侧DLPFC(M7,M13)、双侧PAR到同侧MFC(M9,M15)的网络连接的超越概率大于5%,其中左侧PAR到左侧MFC(M9)的连接的超越概率最大,其他连接超越概率近似等于零。Elderly组:双侧PAR到同侧DLPFC(M7,M13)、双侧PAR到同侧MFC(M9,M15)、左侧DLPFC到左侧MFC(M3)、左侧DLPFC到左侧PAR(M8)、左侧MFC到左侧PAR(M10)和右侧PAR到左侧PAR(M12)的连接超越概率大于5%,其中右侧PAR到右侧DLPFC(M13)的连接的超越概率最大,其他连接的超越概率近似等于零。Youth组:双侧PAR到同侧DLPFC(M7,M13)、双侧PAR到同侧MFC(M9,M15)的网络连接的超越概率大于5%,其中左侧PAR到左侧MFC(M9)的连接的超越概率最大,其他连接超越概率近似等于零。MCI组有效连接的总超越概率显著小于Elderly组(P0.05)和Youth组(P0.01)。(2)不同方向连接族的超越概率MCI组、Elderly组和Youth组的正向连接族超越概率(EP_forward)均显著高于同组的反向连接族(EP_backward)和横向连接族的超越概率(EP_lateral)(P0.01);MCI组的EP_forward(0.6113±0.0193)显著高于Elderly组的EP_forward(0.4626±0.0492)(P0.01);MCI组的EP_backward(0.1942±0.0085)显著低于Elderly组的EP_backward(0.2926±0.0440)(P0.01);MCI组和Elderly的EP_lateral之间没有显著性差异。MCI组的EP_forward、EP_backward和EP_lateral和Youth组均没有显著性差异。(3)左右侧连接的超越概率MCI组和Elderly的左侧连接族超越概率(EP_left)和同组右侧连接族的超越概率(EP_right)之间均无显著性差异,Youth组的EP_left显著大于EP_right(P0.001)。MCI的EP_left(0.5671±0.0751)和Elderly组的EP_left(0.5259±0.0592)之间没有显著性差异;MCI组和Elderly组的EP_left均显著小于Youth组。研究结论:1.行为学结果表明,MCI组工作记忆任务的平均反应时间显著增加,正确率显著降低。工作记忆EEGs时频分析结果表明,工作记忆EEGs的特征频段是theta频段,特征时段是记忆延迟阶段。MCI患者在工作记忆EEGs特征频段(theta频段)的活动减弱。2.源分析结果表明,MCI组的工作记忆激活脑区为MFC,Elderly组的激活脑区为DLPFC和MFC,Youth组的激活脑区为DLPFC、MFC和PAR。MCI组工作记忆的激活脑区比Elderly组和Youth组减少。3.DCM分析结果表明,MCI组的工作记忆有效连接在网络中的贡献程度低于Elderly组和Youth组的有效连接在网络中的贡献程度。4.MCI组和Youth组的工作记忆延迟阶段信息主要流向为从顶叶皮层到前额叶皮层;Elderly组工作记忆延迟期的信息流向既有从顶叶皮层到前额叶皮层,还有从前额叶皮层到顶叶皮层。提示正常老年人不存在工作记忆障碍可能与其工作记忆信息流动增加有关,而MCI患者的工作记忆脑网络没有形成这种代偿机制。
[Abstract]:Research objectives: Mild Cognitive Impairment (MCI) is a cognitive disorder syndrome between normal aging and Alzheimer 's Disease, AD (AD). Working memory disorder is the main clinical manifestation of MCI. The study of the mechanism of working memory disorder in MCI is of great significance. This paper is a hot spot in the field of neuroscience. From the point of view of brain network, this paper makes a dynamic causal model (Dynamic Causal Modeling, DCM) analysis of MCI patients' multi-channel electroencephalographs (EEGs) data under the task memory task, and studies the connection mode of the working memory brain effect network of MCI patients, in order to explore M. The mechanism of CI's working memory disorder provides a new perspective. 1. subjects and experimental data: group MCI, normal old age group (group Elderly) and 15 normal youth group (group Youth) each, the behavioral data and.2. behavioral analysis of 34 channel EEG data of each subject were recorded respectively in the visual working memory task: MCI group, E The average reaction time and the correct rate of.3. data preprocessing under the working memory task of group lderly and Youth: removal of frequency interference, baseline drift, eye movement interference and electromyography interference.4. time frequency analysis for collected EEGs: using the short time Fourier transform (STFT) method to calculate the energy density distribution of the pretreated EEGs and analyze the working memory EEGs Characteristic frequency and characteristic time period.5. source analysis: using the method of distributed linear model to determine the.6.DCM analysis of MCI group, Elderly group and Youth group in the active brain area of the working memory characteristic period and frequency band: Based on the working memory identified by the source analysis, the brain area is activated and the DCM network is constructed, and the transcendental probability of the DCM network subconnection of three groups of subjects is calculated. Exceedance Probability (EP), analyze the connection characteristics of the working memory brain effect network, and calculate the transcendental probability of the three groups of subjects in the DCM network (positive connection, reverse connection and lateral connection) and the left and right connections in the DCM network, and compare the transmission direction of the working memory information of the MCI group and the statistics of the cerebral laterality.7.. Study analysis: the experimental data were statistically tested by single factor ANOVA on SPSS 22 statistical analysis software. The results were as follows: the average reaction time of working memory tasks in group MCI, Elderly group and Youth group were 1387.8 + 80.0ms, 1262.1 + 69.9ms, 912.7 + 66.4ms, and the reaction time of MCI group was significantly greater than El The correct rates of group derly (P0.01) and group Youth (P0.001) were 73.38 + 1.10%, 82.58 + 1.42% and 92.53 + 1.49%, respectively. The correct rate of MCI group was significantly lower than that of Elderly group (P0.01) and Youth group (P0.001). 8Hz), in group MCI, the energy density of EEGs in the theta band was significantly lower than that in group Elderly and Youth. The time frequency diagram of the working memory characteristic channel of the three groups showed that the energy density of the working memory delay period was the most continuous and the concentrated.3. working memory activated brain region source analysis showed that the activation brain area of MCI group was bilateral middle frontal cortex (middle). Line-frontal cortex, MFC); the activation brain area of group Elderly is bilateral dorsolateral prefrontal cortex (dorsolateral prefrontal cortex, DLPFC) and bilateral MFC. The activation brain region of the Youth group is bilateral DLPFC, bilateral MFC and bilateral parietal cortex (bilateral parietal cortex). (1) the transcendental probability MCI group of DCM network subconnections: bilateral PAR to ipsilateral DLPFC (M7, M13), bilateral PAR to ipsilateral MFC (M9, M15) network connection transcendental probability is greater than 5%, of which the left PAR to the left MFC is the greatest transcendental probability, and the other connection transcendental probability is equal to zero. R to the same side MFC (M9, M15), the left DLPFC to the left MFC (M3), the left DLPFC to the left PAR (M8), the left MFC to the left and the left, to the left, the transcendental probability is greater than 5%, and the transcendental probability of the connection of the right side to the right is maximum, and the transcendental probability of the other connections is approximately equal to that of the same side. LPFC (M7, M13), the transcendental probability of the network connection between the two sides PAR to the ipsilateral MFC (M9, M15) is greater than 5%, of which the left PAR to the left MFC (M9) is the greatest transcendental probability, and the total transcendental probability of the other connection transcendental probability approximated to the zero.MCI group is significantly smaller than that of the Elderly group and the group. (2) the connection family in different directions. The more probability of MCI group, the positive connection group transcendental probability (EP_forward) of Elderly group and Youth group was significantly higher than that of the same group, EP_backward and EP_lateral (P0.01), and EP_forward (0.6113 + 0.0193) of MCI group was significantly higher than EP_forward (0.4626 + 0.0492) of Elderly group (P0.01). 0.1942 + 0.0085) was significantly lower than that of group Elderly (0.2926 + 0.0440) (P0.01), and there was no significant difference in.MCI group EP_forward between MCI and Elderly EP_lateral. There was no significant difference between EP_backward and EP_lateral and Youth groups. (3) the overvietnam probability MCI group and the left side connective surpassing probability of the left and right connections (3) There was no significant difference between the transcendental probability (EP_right) of the same group on the right side. The EP_left of group Youth was significantly greater than that of EP_right (P0.001).MCI EP_left (0.5671 + 0.0751) and EP_left (0.5259 + 0.0592) in Elderly group. The EP_left of MCI and Elderly group was significantly smaller than that of the group. Conclusion: 1. behavior science The results showed that the average response time of the working memory task in the MCI group was significantly increased and the correct rate was significantly reduced. The EEGs frequency analysis of working memory EEGs showed that the frequency band of working memory EEGs was the theta band, and the characteristic period of the memory delayed stage was that the activity of.MCI in the working memory EEGs frequency band (theta band) weakened the.2. source analysis node. The results showed that the working memory activated brain area of group MCI was MFC, the activation brain area of group Elderly was DLPFC and MFC, and the activation brain area of group Youth was DLPFC, MFC and PAR.MCI group working memory activated brain regions decreased.3.DCM analysis results than Elderly and Youth groups. The contribution of the Youth group to the network, the information of the working memory delay phase of the group.4.MCI and the Youth group mainly flows from the parietal cortex to the prefrontal cortex, and the information of the Elderly group's working memory delay time flows from the parietal cortex to the prefrontal cortex, as well as the frontal cortex to the top leaf cortex. No working memory impairment may be associated with increased information flow in working memory, while working memory brain networks in MCI patients do not form this compensatory mechanism.
【学位授予单位】：天津医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R749.1;TN911.7

【参考文献】