NMPC for propofol drug dosing during anesthesia induction

Abstract

This paper presents the application of nonlinear predictive control to drug dosing during anesthesia in prospective patients for undergoing surgery. A single-input (propofol) single output (Bispectral index (BIS)) patient model has been employed. The pharmacokinetic-pharmacodynamic model, which is in fact a Wiener-type model, has been used for prediction. A set of 12 patient models were studied while controlling BIS at 50 by applying our in-house nonlinear extended-prediction self-adaptive control strategy (NEPSAC). The results of this simulation study show that NEPSAC outperforms EPSAC, using a nominal patient model for prediction. © 2009 Springer Berlin Heidelberg.