0000000000303286
AUTHOR
Georg Thiele
On the improved Estimation of the Compliance-Parameters of the Physiologically Closed Cardiovascular System
The influence of the type of the sampling-schedule and of the system-excitation with respect to identifiability and estimation-accuracy of the compliance-parameters of a nonlinear model of the physiologically closed cardiovascular system will be investigated. Using the sensitivity-functions of the arterial systemic pressure with respect to the compliance-parameters it will be shown how to select the sample-schedule and system-excitation in order to get optimal estimation-accuracy.
Evaluation of Hemodynamic Parameters for Adaptive Control of the Artificial Heart by Simulation of the Vascular System
Abstract A nonlinear dynamic mathematical model of the cardiovascular system is outlined, which is employable to analyse the hemodynamic system behaviour under normal as well as under pathological conditions. By modifying this model, assuming that cardiac output is constant at a preset level (which is in accordance with most of the total artificial heart systems, since their pumps output is preselected with fixed driving parameters), the hemodynamic behaviour of a circulation system with an artificial heart was simulated. Comparing the simulation results with those obtained from total artificial heart experiments, a good agreement in right atrial and aortic pressure under exercise condition…
Simulation and Parameter-Identification of the Closed-Loop Cardiovascular System by the Use of a Nonlinear Mathematical Model
Abstract A non-linear model of human cardio-vascular system, which includes the short-term pressure regulation mechanism was mathematically derived and implemented on a PDP-11/45 in a block-oriented, interactive programming language. In this way the behaviour of the arterial and venous pressures, cardiac output, stroke volume, total peripheral resistance and heart-rate under ergometric workload was studied by simulation, and experimentally proved, that the model matches the system with good accuracy. In a second phase the parameters of the model were identified by the use of the output-error method. For this purpose a non-linear system identification programming package was applied. The par…
Reliability of Parameter Estimation Methods Applied to the Identification of Biomedical Multicompartment Systems
Abstract The reliability of parameter estimation in biomedical multicompartment models is influenced by the chosen mathematical representation of the model, by the sensitivity of the output on the relevant parameters to be identified and by the experimental conditions, e.g. choice of the data-set. this paper the reliability of different parameter estimation methods applied in a multi-compartment model of the circulatory system has been studied under the assumption that there exists a partially system output insensitivity against the compliance parameters. Suitable conditions for efficient parameter estimation have been found by simulation of sensitivity equations of the system, which is of …