Semiadaptive Infusion Control of Medications With Excitatory Dose-Dependent Effects

This brief presents a closed-loop control approach to infusion of medications that exhibit excitatory dose-dependent effects. A unique challenge associated with closed-loop control of such medications is that the upper limit of the medication-induced excitatory response is unknown, presenting a severe challenge in estimating the parameters in the traditional dose-response model. To address this challenge, we proposed a new dose-response model and semiadaptive (SA) control approach applicable to the closed-loop infusion control of excitatory medications. The new dose-response model eliminates the need for a priori knowledge of the upper limit of the medication-induced response via a novel parameterization to capture local dose-response relationship from the baseline to a target set point and a nonlinear function to convert the depressive response to an excitatory response. The SA control approach makes it possible to apply the well-established MRAC technique to the new dose-response model via selective adaptation of high-sensitivity parameters. We examined the efficacy of the proposed approach using an example of heart rate response to a vasoactive medication norepinephrine. System identification analysis using experimental data and in-silico controller testing suggested that the new dose-response model could faithfully reproduce the experimental data, and that the SA controller could effectively regulate the response in a wide range of simulated subjects.