The Study On Nonlinear Dynamic Characteristics Of EEG Signals Based On Entropy In Mild Cognitive Impairment

Mild Cognitive Impairment(MCI)as the characteristic feature of early stage of Alzheimer’s Disease(AD),has become the focus of the current early diagnosis of AD.Diabetes mellitus is an important factor leading to the decline of cognitive function in the elderly.Therefore,it is of great significance to study the correlation between EEG signals and cognitive function in diabetic patients.The article analyzes the nonlinear characteristics of MCI EEG signals in diabetic patients by investigating mutual entropy and multivariate multiscale sample entropy methods.Firstly,the basic theory of EEG signal is briefly expounded.The neuropsychological scale for evaluating MCI is introduced,and the method of EEG signals acquisition and the preprocessing method are also described.Secondly,the cross-sample entropy,the cross-fuzzy entropy and the cross-fuzzy measure entropy are studied.The coupled coefficients,data length,and threshold r of the three algorithms were calculated by coupled Gaussian and coupled MIX(p)models.The simulation experiments,indicate that the stability and anti-noise performance of the cross-fuzzy measure entropy algorithm performs the best.Therefore,the cross-fuzzy measure entropy algorithm is applied to real EEG signals of diabetic patients with or without MCI,and there is a significant difference in the cross-fuzzy measure entropy values between the MCI group and the control group.The Pearson linear correlation analysis between the cross fuzzy measure entropy values and cognitive function shows that there is a significant correlation between them.Finally,multiscale sample entropy and multivariate multiscale sample entropy are studied and the characteristics of the two algorithms are analyzed by random white noise signal and 1/f noise signal.Simulation analysis shows that the multivariate multiscale sample entropy algorithm can reflect the autocorrelation within the EEG signals channel and the long-range correlation between the channels,thereby quantifying the complexity of the EEG signal.Then multivariate multiscale sample entropy algorithm is applied to analyze the complexity of EEG signals in patients with mild cognitive impairment.It is found that there is a significant difference in the complexity of EEG signals between the two groups of subjects,and the Pearson linear correlation analysis is applied to detect the correlation between complexity and cognitive function.