Automatic control of the extra-corporal bypass: system analysis, modelling and evaluation of different control modes.

Automatic control of the blood gas parameters during extracorporeal circulation has the potential to improve the quality of this procedure and to relieve the personnel from a time consuming task. This paper describes a model of the underlying system for a standard clinical set-up and pinpoints the major difficulties which are the variations of the process gains and the blood- and gas-flow dependent dead times and time constants. Scheduled PI-controllers both for the arterial oxygen as well as for the carbon dioxide partial pressure were designed. Scheduling was based on the blood flow rate. These controllers were tested in a simulation environment. The control systems remained stable under all tested operating condition, but if the blood flow rate was changed abruptly rather large load errors occurred. The performance was improved markedly by adding a feed-forward control path which directly influences the actuating signals based on the actual blood flow rate and the hemoglobin contents, variables which are measured anyway. The major conclusion of this study is to use such direct feed-forward compensation even if more sophisticated control algorithms are used.

Automatic control of isolated organ perfusion.

If provided with adequate physiological conditions explanted hearts may continue to operate in the regular beating mode. This property offers the opportunity to substitute extensive animal experiments by investigations on the isolated organs. At the same time it enables the development of an alternative method for the transport of donor hearts for transplantation. Experimental setups for both applications are described in this paper. The focus is laid on the optimisation of handling and effectiveness of these setups by means of automatic control. The method for investigating the controlled system ("plant") for blood gas exchange, and especially the determination of the transfer function for the partial pressure of oxygen are described. The plant was triggered with stepwise changes on the gas side at several operating points. A first order lag element with time delay was chosen as approximation of every transfer function. The parameter 'time delay', 'gain of the plant' and 'time constant' were analysed as to the dependency on blood flow rate, gas flow rate and partial pressure of oxygen at the blood outlet of the oxygenator. As a result an equation was found to calculate the time delay from gas flow rate and blood flow rate. Correlation of gain and time constant with parameters of the plant were obtained, too. The data is used for the design of a controller, adapting to the different operating points of the plant.

[Biventricular defibrillation using transvenous electrodes]. / Untersuchungen zur biventrikulären Defibrillation mittels transvenöser Elektroden.

This study investigated the feasibility of transvenous biventricular defibrillation using an electrode in a left ventricular vein. The standard lead configuration with a coil in the right ventricle (RV) and a coil in the superior vena cava (SVC) was compared with an additional unipolar coil in an accessible epicardial vein. Only biphasic shocks were used with different shocking modes between the coils in the RV, LV, SVC and the ICD-generator (CAN). Shocks were applied starting with 30 J, decreasing till the DFT was reached. As a result there is a lower DFT when defibrillation is performed including the left ventricular electrode. The impedance did not show a significant increase after more then 20 consecutive shocks. It is a feasible and workable application that might help to reduce the energy demand and increase the safety of such a system.

[Blood gas regulation in isolated beating pig hearts]. / Blutgasregelung bei isoliert schlagenden Schweineherzen.

Continuous perfusion with oxygenated blood is required to supply isolated beating pig hearts. To meet physiological conditions the blood gas parameters pH, pO2 and pCO2 should correspond with values usually found in pigs, respectively. Controlling these parameters manually is a very time consuming task and is therefore done best by a control system consisting of gas blender, oxygenator and blood gas sensors. As this control system should be used with an transportable perfusion apparatus, this paper describes an approach where the gassing is realized only with air and oxygen. In this approach pO2 and pH are measured on blood side so that the control system can change oxygen concentration and gas flow adequately on the gas side.