Electroencephalogram Analysis Based On Multiscale Mutual Mode Entropy And Phase Slope Index

In quest of the effective signal processing techniques to extract the brain waves’ physiological and pathological information,this article adopts the algorithms called Multiscale Mutual Mode Entropy and Phase Slope Index to respectively anlayze complexity and information flow’s direction of epileptic electroencephalogram(EEG)and related attention EEG.This paper discusses the characteristics of epileptic EEG and related attention EEG using Multiscale Mutual Mode Entropy.Results show that the healthy’s average entropy value is higher than epileptic’s value in analysis of original signals and α,β,θ,δ rhythm signals.There exist large differences between the two kinds of people in larger scale analysis of original signals,and θ,δrhythm signals.However,the difference in analysis of α and β EEG is less obvious.In three kinds of attention status,the entropy results show a significantly different change trend with the change of scale.In larger scale,average entropy value of the original EEG and α,β,θ EEG in close-eyes status is significantly higher.The original signals’ and β signals’ average entropy value of count status is slightly higher than that of idle status.In count status,α signals’ average entropy value is close to the results in idle status,and θ signals’ results is lower than that in idle status.This paper discusses information flow’s direction of the epileptic EEG and related attention EEG using Phase Slope Index algorithm.The results show that α rhythm information flow’s average direction of healthy people is from EEG channel to electrocardiograph(ECG)channel and the epileptic have the reverse direction.Their average direction of β rhythm information flow are both from ECG channel to EEG channel and δ rhythm information flow both from EEG channel to ECG channel.In a series of spectrum analysis,the results of two groups in some leads(C4,F4,P4,T3)exist significant differences in information flow’s direction.The information flow’s direction is from front to back in some attention status.Besides,the results of different attention status in some leads(T4,T6,O2,T3,Fp1)exist significant differences in information flow’s direction.Multiscale Mutual Mode Entropy can well quantify the coupling degree between two simultaneous acquisitions of EEG time series on different scales.Phase Slope Index is good to estimate information flow’s direction between two channels in specific frequency band.As an indicator of measuring brain function,the two algorithms can also come to the aid of clinical pathological assessment,brain disease diagnosis,monitoring and testing in brain function recovery.