A Model Predictive Control with Fault Tolerance Concept to Regulate Hypnosis during Anesthesia

Abstract

Anesthesia automation preferred BIS (bi-spectral index derived from electroencephalogram) signal to regulate hypnotic agent propofol infusion during surgery to avoid adverse reactions and to reduce post recovery time. But BIS signal may be suspend during surgery due to the poor quality of EEG signal. Thus, fault tolerance ability against BIS suspension is essential to prevent undesirable states such as intraoperative arousal and over/under infusion of propofol. This paper proposes a model predictive control (MPC) with multivariable fault tolerance concept based design to regulate propofol dose. The main aim is to design robust system with fault tolerance ability that works efficiently to maintain smooth BIS with less variations of propofol dose during normal and BIS sensor failure condition. In this paper sensor fault and hemodynamic surgical disturbance is considered to evaluate the performance of control system. Real pharmacological data of 5 different patients are used in proposed design.