Implementation of Model Predictive Control for Closed Loop Control of Anesthesia

Abstract

This paper focuses on the design and implementation of model- predictive controller (MPC) for the close loop control of anesthesia for a patient undergoing surgery. A single input (Propofol infusion rate) single output (bispectral index (BIS)) model of patient has been assumed which includes variable dead time caused by measurement of bispectral index. The main motivation for the use of MPC in this case relies on its ability in considering, in a straightforward way, control and state constraints that naturally arise in practical problems. The MPC can take into account constraints on drug delivery rates and state of the patient but requires solving an optimization problem at regular time intervals. We proposed a modified active set method algorithm using Cholesky factorization based linear solver to solve online convex quadratic programming (QP) problem, to reduce complexity of the algorithm, eventually to accelerate MPC implementation. Experimentation shows that excellent regulation of bispectral index (BIS) is achieved around set-point targets.