Nonlinear Model Predictive Control of the Depth of Hypnosis during Anesthesia in the Presence of Uncertainty

In this paper, a robust nonlinear model predictive control (NMPC) approach is proposed to control the depth of hypnosis by taking into account the uncertainty about patient’s pharmacodynamic parameters related to the influence of propofol on BIS (the bispectral index). The novelty of the approach consists in that the robust optimal control problem is based directly on the nonlinear dynamic model describing the drug effect on BIS and it explicitly considers the presence of time delay in the response of BIS to drug infusion. The approach is applied for computing the BIS trajectories of 24 patients in the nominal case, as well as for the worst-case scenario about the uncertainty. The obtained results show that the robust NMPC controller brings BIS to the required interval and close to the set-point value despite of the parameter uncertainty and the time delay.