Modelling of Cardiovascular Regulation in Humans

Abstract

<p> A linear state space model (LSSM) of cardiovascular regulation in ten normal human volunteers is developed using instantaneous lung volume (IL V), heart rate (HR.), pulse pressure (PP) and mean arterial blood pressure (MBP) time series. Closed-loop transfer functions are computed and physiologically interpreted and the sensitivity of the transfer functions is assessed by comparison of supine and standing experimental results. The zeros ofthe transfer functions are used to infer the causality relationship between HR and PP. Results (1) In the supine condition, changes in ILV cause changes in HR within 0.5 s, followed shortly (0.3 -0.5 s) by changes in PP and finally changes in MBP 1 -2 slater. (2) When standing, changes in MBP occur concurrent with changes in PP. (3) MBP changes are dominated by blood pooling effects when standing. ( 4) Group delay is less affected than the magnitude by the physiological differences between the supine and standing conditions. (5) The relationship between HR and PP is neither completely causal nor anti-causal, but rather a combination of the two. (6) The minimum system delays are coincident with breathing frequencies between 0.2 -0.4 Hz. Conclusions Closed-loop LSSM and transfer function analysis may be used to infer the time delays and causality of the closed-loop system response. The classical model of RSA generation is supported by the LS SM results. </p>