Research On The Synchronization And The Nonlinearity Analysis Algorithm Of Eeg Signals With Mild Cognitive Impairment

With the aggravation of population aging in China,the prevalence and incidence of Alzheimer’s disease are increasing rapidly and continuously.Mild cognitive impairment(MCI)is the early stage of Alzheimer’s disease.In-depth analysis of MCI electroencephalograph(EEG)signals to find rapid and effective EEG technical indicators to assist the evaluation and diagnosis of MCI and prevent the deterioration of the disease.It is of great significance to the public health security of our country,and it is also a hot spot in the field of brain science.Synchronization and nonlinearity are important features to characterize MCI EEG signals.Based on the recursion principle,this dissertation proposed four new EEG analysis algorithms.These algorithms were used to analyze the resting EEG signals of cognitively normal controls and MCI patients collected by the Chinese People’s Liberation Army Rocket Force Hospital.The differences of EEG synchronization characteristics and nonlinear characteristics on multiple spatiotemporal scales between the two groups were analyzed from multiple perspectives.These provide a theoretical basis for mining effective EEG technical indicators to evaluate MCI.These have important theoretical significance and application prospects for the diagnosis,prevention,and treatment of MCI.Firstly,the Order Recurrence Synchronization(ORS)algorithm is proposed and used to analyze the synchronization of pairwise EEG signals in different brain regions of the cognitively normal control group and the MCI group.This algorithm does not need to set a threshold and calculate the distance between state vectors.Simulation results show that the algorithm is less affected by parameters and has better anti-noise performance.The results of EEG signal analysis show that the greater the spatial distance of EEG channel pairs,the lower the synchronization value.The MCI group has lower EEG synchronization values in most channel pairs than the control group,especially in the right cerebral hemisphere.The decrease of EEG synchronization values in P3-P4,O2-T4,and O2-T6 in the MCI group is associated with decline of cognitive and memory function.Secondly,the Global Recurrence Synchronization(GRS)algorithm is proposed,which can effectively quantify the global synchronization intensity of multi-channel EEG signals in the MCI group and the control group.This algorithm uses the ORS method to calculate the correlation values between pairwise channels combines the proxy data method to reduce the influence of random correlation,and can describe the real global synchronization intensity of multi-channel EEG signals more accurately.Compared with the global synchronization index and S-estimator,this algorithm is less affected by signal frequency band and data length,is more sensitive to the change of signal coupling coefficient,and has better performance.The results of EEG signal analysis show that GRS values in the MCI group are always lower than those in the control group,especially in the parietal,occipital,and right temporal regions,and are associated with the decline of cognitive function,attention,and executive ability.Then,the Multiscale Dispersion Recurrence Plot(MDRP)algorithm is proposed,which can analyze the amplitude information of EEG signal,and overcome the disadvantage of the serious loss of signal amplitude information in the ordering recurrence calculation.The simulation results show that this algorithm can reflect the nonlinear characteristics of different state signals under shorter data lengths,is less affected by parameters,and has better anti-noise performance.Applied this algorithm to study the multi-scale determinism(DET)of EEG signals in the MCI group and the control group.The results show that the larger the scale factor,the lower the EEG determinism value.The short-scale DET values of Fz,F3,C3,C4,Pz,and P3 channels in the MCI group are significantly higher than those in the control group and are associated with cognitive decline.Finally,the Multi-variant Multiscale Dispersion Recurrence Plot(MvMDRP)algorithm is proposed and used to analyze the multi-scale recurrence features of multichannel EEG signals in the MCI group and the control group.Lorenz and R?ssler models are used to generate simulation signals and study the effects of scale factor,class number,noise intensity,and data length on this algorithm.The simulation results show that this algorithm can quantify the multi-scale recurrence features of the system at a relatively short data length,effectively reflect the details of its internal structure,and explore the micro and macro variations of multivariate nonlinear time series from a global perspective.The EEG signal analysis results show that the larger the scale factor,the lower the EEG determinism,mean diagonal line length,and recurrence entropy values.The values of short-scale determinism,mean diagonal line length,and recurrence entropy in the central regions and left temporal region are significantly higher in the MCI group than in the control group,which are associated with declines in cognitive and language function.