Model dependency of multivariate autoregressive spectral analysis

摘要: A combination of simulations and experimental data analysis has been used to demonstrate that, because cardiovascular control represents a complex linking input output parameters, interpreting the variability individual parameters such as heart rate arterial pressure virtually requires use techniques that quantify by relating these inputs outputs. Transfer functions represent appropriate for this purpose. Further, despite complexities in vivo physiological control, many elements can be well characterized only taking into account single outputs using bivariate AR model. However, occasionally when two systems have strong simultaneous influence on parameter, respiratory activity RR interval, an expansion model general multivariate case may required complete interpretation. Finally, although not fully demonstrated here, closed-loop nature it is likely algorithms include causality characteristic, formulation, will most accurately identify transfer relations.