Entropy And Recurrence Plot Based EEG Signal Analysis In Mild Cognitive Impairment

Mild Cognitive Impairment(MCI)is a neurodegenerative disease related to memory function decline,which can easily develop into Alzheimer’s disease.In this paper,the recurrence quantitative analysis method of refined composite multivariate multiscale fuzzy entropy algorithm,multiscale dispersion recursive plot and dispersion cross recursive plot is used to study and analyze the nonlinear characteristics of MCI patients’ EEG signals.First,it briefly introduces the generation mechanism and characteristics of EEG signals,and introduces the evaluation standard neuropsychological scale of MCI and the inclusion and exclusion criteria of MCI group and control group.Secondly,the refined composite multivariate multiscale fuzzy entropy algorithm uses the coupled MIX(p)model,1/f noise and white noise combined model for simulation analysis,and the algorithm is compared with three multichannel entropy algorithms.It is found that the fine compound multi-scale fuzzy entropy algorithm is more suitable for analyzing the MCI patients’ EEG signals.The MCI EEG patients’ signals were analyzed by using refined composite multivariate multiscale fuzzy entropy,and it was found that the MCI group and the control group showed significant differences in the long-scale.By using Pearson’s linear correlation analysis,it is found that there is a significant correlation between the refined composite multivariate multiscale fuzzy entropy of EEG signals and neuropsychological test scores.After that,the multiscale multichannel synchronization value of each brain area was briefly explored,and the multiscale normalized sorting mutual information global synchronization index was used for analysis.The study of multiscale global synchronization can provide some references for the research and selection of multichannel entropy algorithms.Then,the obtained dispersion recursive plot algorithm is simulated and analyzed using the Logistic model,and the method is compared with the Order recursive plot algorithm,and it is found that the performance of the dispersion recursive plot is better than that of the Order recursive plot.And combining the dispersion recursive plot algorithm with recursive quantitative analysis to study MCI patients’ EEG signals,most electrodes show significant differences at large scales,and there is a significant correlation with neuropsychological test scores.Finally,the dispersion recursive plot is extended to the dispersion cross recursive plot,and the dispersion cross recursive plot algorithm is combined with recursive quantitative analysis,and the dual-channel coupling MIX(p)model is used for simulation analysis.It is found that the algorithm has low sensitivity to data length and noise.Using dispersion cross-recursive plot to study the cross-determinism of electrode pairs in different brain regions,it is found that there are significant differences between the MCI group and the control group,and there is a significant correlation between the certainty value and neuropsychological test scores.