Coupling Analysis Of Short-term ECG Time Interval Series And Blood Pressure Series

Background:The autonomic nervous system(ANS)plays a key role in the physiological regulation of cardiovascular system.The imbalance of ANS may cause damage to cardiovascular function,resulting in the occurrence and development of various cardiovascular diseases.Time interval signals from ECG and arterial blood pressure are under control of ANS and present certain dynamic characteristics such as heart rate variability(HRV)and blood pressure variability(BPV).The dynamic interactions of these physiological signals contain important information about autonomic nervous regulation.Based on the time interval series from ECG and systolic blood pressure(SBP)series,this study is to perform the joint analysis between RR interval(RRI)and SBP as well as RRI and QT interval(QTI)to investigate the non-invasive evaluation method of autonomic nervous state by characterizing the casual interactions of RRI-SBP and RRI-QTI.Methods:The data for the joint analysis of RRI-SBP were from database PRCP(n=10)provided by PhysioNet.This part is aimed to investigate the changes of the coupling strength between RRI and SBP before and after head-up tilt(HUT)with different tilt speeds from dynamic and steady perspectives.Beat-by-beat time series of RRI and SBP were extracted from both slow tilt(ST,75°HUT over 50 s)and rapid tilt(RT,75°HUT over 2 s).Then,time-frequency analysis and information decomposition analysis,combined with time-domain indexes and short-term fractal exponent(al)were applied to perform joint analysis between RRI and SBP.The data for the joint analysis of RRI-QTI were from database Normal(control group,n=189)and database ESRD(pathological group,n=43)provided by the Telemetric and Holter ECG Warehouse(THEW).This part is aimed to explore the method of characterizing the behavior of autonomic nervous based on circadian rhythm under physiological and pathological conditions.A causal cross-spectral approach based on the autoregressive model was applied to analyze the linear coupling from RRI to QTI.Result:For the analysis of RRI-SBP,the results of information decomposition analysis indicated that all of the significant differences appeared in the feedback direction(SBP?RRI)due to baroreflex control on RRI.The prediction of RRI after ST significantly increased compared to that in supine position(0.416±0.067 vs 0.626±0.127),indicating the elevation of the coupling strength along the baroreflex.However,HUT showed few effects in the feedforward direction of RRI?SBP.There were no significant differences between ST and RT for all of the same indexes before HUT.However,the coefficient of variation of RRI(CVRRI)in the steady state after RT was significantly increased and ?1 was significantly decreased compared to that after ST despite the fact that there was no difference for RRI.What's more,the results of time-frequency analysis suggested the different behavior of dynamic response to ST and RT.Our research proved the effectiveness of information decomposition analysis to detect the dominant causal direction(feedback or feedforward)in the RRI-SBP interactions and to characterize the changes of the prediction of RRI and SBP signals before and after HUT.For the analysis of RRI-QTI,the results of causal cross-spectral analysis indicated that for Normal,there were significant differences between daytime and nigh-time results of GLF,GHF and LF/HF.However,no significant diurnal variation of the corresponding indices existed in ESRD;GLF-d/GLF-n and(LF/HF)d/(LF/HF)n were significantly smaller in ESRD than that in Normal while there was no statistic difference of GHF-d/GHF-n between Normal and ESRD.The above results revealed that the loss of circadian rhythm in ESRD was primarily caused by the overactivity of sympathetic branch at night compared to Normal.Conclusions:In this study,the joint analyses of short-term RRI-SBP and RRI-QTI were performed to explore the methods of evaluating the behavior of cardiac autonomic nervous based on the dynamic time interval series from ECG and beat-by-beat blood pressure series.Our proposed methods are expected to provide a promising method on ANS evaluation for clinical application.