| [Objective]This paper aims to make research on hypersynchrony discharging system ofepileptic attack by making synchronism analysis of EEG phase of epilepsy focus.[Methods]This paper carries on phase synchronization analysis on 4(C3, P3, O1, F7)EEGof the same group of the sample by application of cross correlation function, mutualinformation and synchrony index. Collect EEG signal of temporal-lobe epilepsypatients and that of normal people at the standard N=7680 and f_s=256HZ. Andapply Hilbert to count the phase position of each EEG signal and carry on analysisand comparison on synchronicity of the EEG phase position of this area byapplication of cross correlation function, mutual information and synchrony index.Selection of the length of dissolution of Mexican Hat successive wavelettransform is decided by the length and sampling frequency of the data of collectedEEG signal. In this paper, N=7680, f_s=256HZ and the maxim of the length iscounted to be 8 by application of wmaxlev function of Matlab. As in this paper,f_s=256 HZ. It decides that the length selected should be 6. Scope of the disolusionfrequency are: first length: 64-128HZ, second length:32-64HZ, third length:16-32HZ, fourth length: 8-16HZ, fifth length: 4-8HZ, sixth length:0-4HZ which correspond to main frequency scope ofδfrequency band,θfrequency band,αfrequency band andβfrequency band. This paper will makephase synchronization analysis on the above mentioned frequency bands. [Results]1.The result of epilepsy group and normal group:epilepsy group normal groupcross correlation function: 0.8285±0.0573 0.4602±0.0587;mutual information: 0.7758±0.0512 0.3128±0.0569;synchrony index: 0.7006±0.0498 0.3925±0.0522.2. The result of each component of wavelet between epilepsy group and normalgroup:(1)αfrequency band:(p＜0.01)epilepsy group normal groupcross correlation function: 0.8732±0.0621 0.4482±0.0496;mutual information: 0.8022±0.0593 0.3425±0.0489;synchrony index: 0.7832±0.0566 0.7832±0.0566.(2)θfrequency band:(p＜0.05)epilepsy group normal groupcross correlation function: 0.8482±0.0492 0.5628±0.0583;mutual information: 0.7282±0.0465 0.4023±0.0426;synchrony index: 0.7455±0.0621 0.7455±0.0621.3. There is not much difference of phase synchronism betweenδfrequency bandandβfrequency band(p＜0.05).[Conclusions]1. It is found through analysis of phase synchronism of EEG of 8 temporal-lobeepilepsy patients and 8 normal people by application of cross correlation function,mutual information and synchrony index that: whole phase synchrony of EEG signalof epilepsy patients is obvious higher than that of normal people.2. Compared with the three methods, difference of the result got through analysisby mutual information is most obvious. The reason is that mutual information whichis a nonlinear synchronizing analysis method is the most suitable to be applied tomake analysis on EEG signal which is nonlinear. 3. Phase synchronism of a frequency band of epilepsy group is most obvioushigher than that of normal group (P＜0.01) and synchronism of EEG phasecomponent of this frequenvy band is higher than that of the whole EEG. Therefore,αfrequency band is the key to make analysis on hypersynchrony discharging systemof epilepsy EEG.4. As to phase synchronism of each component of wavelet, there is not muchdifference of phase synchronism betweenδfrequency band andβfrequency bandof normal people (P＞0.05). While phase synchronism ofδfrequency band is muchhigher than that ofαandβfrequency bands. |
PDF Full Text Request