Transitions In Physiological Coupling Between Heartbeat And Pulse Across Sleep Stages

In recent years,the proportion of people suffering from sleep disorders has gradually increased.People are increasingly in need of accurate and reliable monitoring of sleep quality technology,and then take appropriate measures to adjust and treat.Today,clinically,the sleep state is monitored by polysomnography PSG using a wide range of sleep apnea signals from subjects.However,PSG operation is very complex and costly,which is not conducive to routine monitoring.Therefore,it is of clinical significance to study the relationship between non-EEG parameters and sleep stage.This paper firstly applies the detrended fluctuation analysis algorithm and the detrended moving average algorithm to calculate the autocorrelation of the two signals in different sleep stages by collecting the ECG interval and pulse interval sequences as the characteristic parameters of sleep stages.The Detrend cross-correlation analysis algorithm and the detrended moving average cross-correlation analysis algorithm calculate the coupling strengths of the ECG signal and the pulse signal in different sleep stages to serve as a basis for identifying different sleep stages;finally,the multiple scale index combinations of the different sleep stages are obtained,which are used as the different SVM feature parameters for classification prediction.The research aspects of experimental innovation are as follows:(1)Detrended Fluctuation Analysis(DFA)and Detrended Moving Average analysis(DMA)were performed for RRI and PPI in different sleep periods.We found that DFA(RRI),DFA(PPI),DMA(RRI)and DMA(PPI)scale indices have the rules of wake> rem> ls> ds in different sleep stages,,and that the range of scale exponents corresponding to each sleep stage is very easy to distinguish.The DFA(RRI),DFA(PPI),DMA(RRI),DMA(PPI)scale index timing diagrams and AASM artificial staging sleep state staging have a high degree of synchronization and the same change rules.(2)Next,we analyzed the coupling intensity of RRI and PPI application of Detrend cross-correlation analysis(DCCA)and Detrend Moving average cross-correlation analysis(DMCA)for different sleep periods.We found that DCCA(RRI-PPI)and DMCA(RRI-PPI)standard index have the rules of wake> rem> ls> ds in different sleep stages,and that the corresponding standard for each sleep stage is very easy to distinguish.The time series of DCCA(RRI-PPI)and DMCA(RRI-PPI)coupling strength and AASM artificial staged sleep status have a high degree of synchronization and the same variation rules.(3)Apply the characteristic parameters of different sleep stages obtained by(1)and(2)in SVM,namely,DFA(RRI),DFA(PPI),DMA(RRI),DMA(PPI),DCCA(RRI-PPI),DMCA(RRI-PPI)was combined into 5 groups.The coupling strength parameters of RRI and PPI signals are compared with the DFA and DMA parameters of a single signal.The results show that DCCA and DMCA have the best classification accuracy,followed by DMA.If DFA and DMA are combined,compared with DFA and DMA respectively,in addition to REM period,WAKE period,LS period,and DS period accuracy are improved;but compared with DCCA and DMCA,in addition to the DS period classification accuracy is improved.Outside,the others are close.Overall,compared with DFA and DMA,the coupling intensity of DCCA and DMCA is best for sleep staging.