Model Predictive Control with Exponential Cost Function to Regulate the Propofol Infusion Rate

Abstract

The anesthesia infusion process has been a topic of study for years due to the dangers that can lead to anesthesia management during general surgery. Therefore, the automation of the anesthesia delivery process is still a relevant problem; the automation problem consists of computing a reasonable infusion rate of the anesthetic drug so that the desired hypnosis level is reached. Most of the approaches are based on a mathematical model of the process, which can be assumed as LTI system with a static nonlinearity in the output. This model considers the drug infusion rate as the input, and the output is the bispectral index (BIS), which measures the hypnosis in the patient. Furthermore, these plant dynamics are based on a compartmental model and consider pharmacokinetics and pharmacodynamics effects. In the present contribution, a description of this mathematical model is presented wherewith a model predictive control (MPC) with exponential cost function is designed to regulate the hypnosis state in the patient. Finally, the proposed algorithm is simulated for a group of virtual patients.