Robustness Tests of a Model Based Predictive Control Strategy for Depth of Anesthesia Regulation in a Propofol to Bispectral Index Framework

This paper verifies the robustness of a model based predictive control scheme for depth of anesthesia (DOA) regulation. The manipulated variable is Propofol, which is used in a Model based Predictive Control (MPC) algorithm for automatic induction and control of DOA. In turn, DOA is evaluated by means of the Bispectral index (BIS). The simulation tests are performed on a set of 17 virtually generated realistic patients with significantly varying sensitivity to Propofol infusion. The results show a high-efficiency, optimal dosage and robustness of the MPC algorithm to induce and maintain the desired BIS reference while rejecting typical disturbances from surgery.