Analysis Of Epileptic Ieeg Signals Based On Cross-frequency Coupling

Cross-frequency coupling(CFC)is the main way of interaction between neural oscillations and an important tool for studying brain function.Among them,phase-amplitude coupling(PAC)and quadratic phase coupling(QPC)have received extensive attention and play an important role in the field of neural information processing.Epilepsy is a brain functional disorder caused by sudden abnormal discharges of brain neurons,which is sudden and recurrent,and some patients have drug resistance.Therefore,the use of i EEG signals to analyze the cross-frequency coupling pattern during epileptic seizures is of great significance for the diagnosis and treatment of epilepsy.First of all,a novel analysis method based on Jensen-Shannon divergence and correlation matrix(JSCM)was proposed to measure the PAC between neural oscillations at different frequencies.The simulation results show that the proposed method can effectively reflect the coupling strength while slightly vary with the data length.This new method outperforms four frequently used algorithms in the robustness of resisting noise and the accuracy of detecting PAC in wide frequency ranges.Secondly,the JSCM method was used to analyze the PAC pattern of i EEG signals of patients with epilepsy during the entire seizure process.Through the single-channel PAC analysis of the i EEG signals at different seizure periods in the seizure onset zone of each patient,and the cross-regional multi-channel PAC analysis of multiple brain regions of each patient,it was found that the PAC showed temporal and spatial specificity in the entire seizure process.In addition,significant differences in the distribution of PAC under different seizure states have been observed.These results provide some references for exploring the mechanism of seizure,transmission and termination.Finally,the bi-phase-locking value(BPLV)method was used to analyze the QPC of i EEG signals in patients with epilepsy.By analyzing the quadratic phase coupling pattern of each patient during the seizure process,it was found that the BPLV has a significant upward trend in about 20 s before the seizure,and the BPLV value has a smaller peak in the middle and late seizures.In addition,the range of interaction frequency and the maximum value of BPLV during epileptic period are larger than those during interictal period,while the coupling frequencies are significantly lower than those during interictal period,which provides a new direction for the study of epilepsy.