A new analytical method (sequential convolution) for describing ventricular-vascular interactions was used to predict instantaneous pressure and flow in four isolated canine left ventricles ejecting into a computer-simulated arterial system. Ventricular pumping ability was described by a load-independent elastance, [E*(t)] combined with a ventricular internal resistance. “Arterial” properties were characterized using a time-based impulse response function that is derived from impedance measurements. Sequential convolution was then used to couple these independent descriptions of ventricular and vascular properties. Predicted pressure-volume trajectories, as well as instantaneous pressures and flows, closely matched the experimental data. Stroke volume, peak pressure, and peak flow were typically within 5% of measured values. This method provides a powerful analytical technique for examining ventricular-vascular interactions and has potential application in evaluating the ventricular-loading effects of more complex in vivo vascular properties.
Strategies for Sequential Prediction of Stationary Time Series
