Dysfunction Mechanism Of Brain Network Connectivity Of Eeg During Working Memory Task In Mild Cognitive Impairment

ObjectiveMild Cognitive Impairment(MCI)is the risk factor of Alzheimer's Disease(AD)and it is also the critical stage of the prevention and treatment of AD.Working Memory Dysfunction(WMD)is the main clinical manifestation of MCI.Studying the mechanism of WMD in MCI has great significance in the prevention and treatment of AD.In this paper,multi-channel electroencephalographs(EEGs)of MCI and normal control under the visual working memory task is used as the research object.Combining behavioral data of working memory,neural oscillations and brain network functional connectivity,the mechanism of dysfunction pattern of brain network connectivity of WMD in MCI is studied to provide support for the quantitative evaluation of WMD in MCI and the prevention and treatment of AD.Methods1.Experimental data: Behavioral data of working memory and 34-channel EEGs were recorded from 15 MCI patients(MCI group)and 15 normal control(Con group)while they performed the visual working memory task.2.Behavioral data of working memory: We calculated the average accuracy and the average reaction time of MCI group and Con group under the visual working memory task to study the behavioral disorders of working memory in MCI quantitatively.3.EEG data preprocessing: We removed 50Hz-frequency interference,baseline drift,electrooculography interference and electromyography interference from the recorded original EEGs of MCI group and Con group.4.Time-frequency analysis: We calculated the time-frequency distribution of the 34-channel preprocessed EEGs of MCI group and Con group under the visual working memory task based on Short-time Fourier Transform(STFT)to analyze the principal frequency band of energy density and the principal period of working memory,and quantitatively evaluated the neural oscillation of MCI in working memory dysfunction.5.Granger causality analysis: We constructed causal network of EEG using Granger causality analysis to determine in which frequency band that the maximum brain network connectivity of EEG in working memory was located.We calculated the brain network functional connectivity of the 34-channel preprocessed EEGs of MCI group and Con group under the visual working memory task based on Directed Transfer Function(DTF)to quantitatively evaluate the brain network connectivity of WMD in MCI.6.Statistical analysis: The difference between MCI group and Con group was tested by independent sample t test,and the paired sample t test was applied to the statistical test of each frequency band in the group.Results1.Behavioral disorders of working memory in MCI: The average accuracy of MCI group and Con group under the visual working memory task were 72.34±0.81% and 83.19±1.47% respectively and the average accuracy of MCI group was significantly lower than the avergae accuracy of Con group(P<0.001).The average reaction time of MCI group and Con group were 1509.4±39.2ms and 1235.6±74.7ms respectively and the average reaction time of MCI group was significantly higher than the average reaction time of Con group(P<0.001).2.Decreased theta oscillation of EEG in MCI:The average energy density of EEGs in theta band(4-8Hz)of both MCI group and Con group under working memory task were the highest.The average energy density of EEGs in theta band of MCI group(0.305±0.033?V^2/Hz)was significantly lower than that of Con group(0.468±0.103?V^2/Hz)(P<0.001).The average energy density of EEGs in retention period was significantly higher than that in encoding period and retrieval period of working memory.The average energy density of EEGs in theta band in retention period of working memory mainly concentrated near Fz channel in frontal midline.3.Decreased brain functional connectivity of gamma network of EEG in MCI: The average network connectivity of EEGs in gamma band(30-80Hz)in retention period of working memory of both MCI group and Con group were the highest.The average network connectivity of EEGs in gamma band of MCI group in retention period of working memory(0.0123±0.0006)was significantly lower than that of Con group(0.0153±0.0005)(P<0.001).The average network connectivity of EEGs in gamma band in retention period of working memory mainly concentrated on parietal lobe.Conclusion1.Behavioral disorders of working memory in MCI :Compared with Con group,the averge accuracy of MCI group is significantly decreased and the average reaction time is significantly increased in working memory task.2.The working memory dysfunction of MCI is closely related to the decreased neural oscillations of theta.Theta band is the principal frequency band of energy density in EEGs of working memory.The principal period of working memory is the retention period.The decrease of average energy density of EEGs in theta band of MCI group is the neural oscillation pattern of WMD in MCI.3.The working memory dysfunction of MCI is closely related to the decreased network connectivity of EEGs in gamma band.Gamma band is the maximum connectivity frequency band in EEGs of retention period in working memory.The decrease of average network connectivity of EEGs in gamma band of MCI group is the network dysfunction pattern of WMD in MCI.