利用Volterra-Wiener模型估算壓力反射敏感度－以EUROBAVAR資料庫為例

Abstract

The short-term control of blood pressure is the arterial baroreflex and the baroreflex sensitivity (BRS) is the indicator that describes such mechanism. The BRS estimation, spontaneous BRS index, is based on beat-to-beat time series analysis of systolic blood pressure and heart rate interval underlying openloop or closed-loop linear model in time or frequency domain. However, the physiological regulatory mechanism of blood pressure is inherently nonlinear and dynamic. It would make a great improvement for accurate assessment of arterial baroreflex by a suitable dynamic nonlinear model. This paper presents a nonlinear Volterra-Wiener model and the Laguerre expansion technique to analyze the changes in systolic arterial blood pressure, as model’s input, and their corresponding changes in heart rate interval, as model ’s output. Compared with the measurements by sequence technique and alpha coefficient technique in EUROBAVAR dataset, we found that the estimates of BRS by the proposed technique were lower either in lying or in standing position and the lying/standing ratio were higher. The proposed technique improves the overestimation of spontaneous BRS index obtained by other approaches such as sequence technique or alpha coefficient technique.