Microprocessor-based controller for pharmacodynamical applications

Abstract

To design an automatic drug delivery system for the purpose of keeping the output as close as possible to a desired value, the system input-output is first represented as an autoregressive model. A minimum variance controller is then designed to solve the control problem. The implementation of the controller is performed on a microprocessor. As a specific application, the control of the mean arterial blood pressure in dogs using a vasodilator drug is presented. An automated drug deliver system is constructed; it contains a microprocessor (Motorola 6800), A/D and D/A converters. The output of the microprocessor drives a pump, which injects the drug at an appropriate rate into the blood circulation of the dog. The algorithms for the microprocessor are based on the on-line parameter estimation and optimal controller equations. The resulting closed-loop system was tested on dogs using nitroprusside as the drug which decreases the blood pressure. The tests were performed over the range of 90-170 mm Hg of the mean arterial blood pressure and over time periods of 6-10 rain. The system constructed maintains the mean arterial blood pressure within ± 10 mHg from the desired constant value. Results from typical experiments are presented and discussed.