Control of neuromuscular blockade with Gaussian process models

This paper presents Nonlinear Model Predictive Control (NMPC) of neuromuscular blockade induced by atracurium on patients subject to general anesthesia. In order to tackle the high levels of uncertainty in the process behavior, probabilistic and non-parametric Gaussian process models are used in the NMPC approach. The proposed control structure was tested in a bank of models that represent patients subject to general anesthesia under elective surgery. All patients models were stabilized and yield a satisfactory performance.

► A Nonlinear Model Predictive Control application to neuromuscular blockade. ► A Gaussian process modeling of neuromuscular blockade behavior. ► The approach shows to be very reliable for a wide range of neuromuscular blockade behaviors and high noise levels.