Search

Abstract Closed-loop anesthesia drug delivery systems can reduce the effect of inter-patient variability, minimize variability in desired clinical effects, and ultimately improve patient safety. Feasibility of closed-loop anesthesia has been clinically evaluated in a number of studies. To obtain regulatory approval to use closed-loop anesthesia as a clinical device, patient safety must be demonstrated. Safety systems for closed-loop anesthesia have been proposed to limit anesthetic concentrations in the plasma and effect-site. Such safety systems minimize the risk of drug under/overdosing. Anesthetic drugs have serious side effects and can significantly reduce blood pressure, thereby jeopardizing patient safety. This paper discusses a safety system that includes safety constraints on blood pressure in addition to the plasma and effect-site concentrations. The proposed safety system is formalized using formal model verification techniques. The performance of closed-loop anesthesia in the presence of constraints on the plasma and effect-site concentrations as well as blood pressure is shown using simulation results. The effectiveness of the proposed safety system is also illustrated in simulation using a realistic clinical scenario. This paper shows that maintaining blood pressure within safety constraints increases the time for anesthesia induction, especially for patients whose blood pressure is more sensitive to anesthetics. For this population, clinical intervention such as administration of vasoactive drugs may be required to counter the effect of propofol on blood pressure during closed-loop anesthesia to achieve a safe yet sufficiently fast induction. This paper extends a previously published formalized safety system which guarantees that the plasma and effect-site concentrations remain within safety limits (Yousefi et al., 2017). Highlights • A formal proof of safety for closed-loop propofol anesthesia is provided. • Closed-loop anesthesia can be performed with minimum risk of drug over/under dosing and hypotension. • Limiting blood pressure increases the induction time of anesthesia. • Administration of vasoactive drugs may be required to have a fast induction of anesthesia while avoiding hypotension. [ABSTRACT FROM AUTHOR]

The analysis in this paper shows that L 1 adaptive state feedback control approximates an implementable LTI controller because: (1) the addition of a low-pass filter in the controller structure leads to an internal loop in the controller and (2) high gain feedback within this loop leads to inversion of the linear reference model and cancellation of non-linear characteristics of the adaptation law. This result holds for a variety of adaptation laws and the achieved performance does not depend on the exact formulation of this law. [ABSTRACT FROM AUTHOR]

This paper studies application of different constrained control concepts for the control of Depth of Hypnosis (DOH) in closed-loop anesthesia to guarantee patient safety, while ensuring acceptable tracking performance. The core idea is to formulate Overdosing (OD) prevention and Blood Pressure Decrease (BPD) prevention as operational constraints, and then use a constrained control scheme to enforce the constraints satisfaction at all times. In this paper, three methods are studied: (1) Explicit Reference Governor (ERG), (2) Safety Preserving Control (SPC), and (3) Model Predictive Control (MPC). The performance of the methods is assessed with respect to a simulated surgical procedure for 44 patients whose models have been identified using clinical data. In particular, three realistic clinical scenarios are studied in this paper: (1) normal mode, (2) stimulation, and (3) low clearance. The results demonstrate cons and pros of each method in closed-loop anesthesia according to clinically relevant assessments. [ABSTRACT FROM AUTHOR]

• Habituating MPC provides a unique control solution for MISO systems. • Extended HMPC provides a clinically relevant MPC control objective function. • Simulation results show that the extended design meets clinical objectives. • This solution can facilitate practical implementation of patient safety constraints. This paper proposes a model predictive control (MPC) solution for closed-loop propofol and remifentanil infusion based on feedback from a depth-of-hypnosis monitor, a multi-input single-output system. A clinically relevant objective function is proposed, extending the habituating MPC solution. The presented MPC controller meets the following clinical objectives; it (1) gives the anesthesiologist control over the analgesic-hypnotic balance, (2) reacts fast in response to disturbances, (3) maintains a bounded response to noise, and (4) is robustly stable toward patient variability. The proposed MPC framework and controller design will facilitate practical implementation of patient safety constraints compared to previously proposed LTI solutions. [ABSTRACT FROM AUTHOR]