Anesthesiologist in the Loop and Predictive Algorithm to Maintain Hypnosis While Mimicking Surgical Disturbance. Issue 1 (July 2017)

• Record Type: Journal Article
• Authors: Ionescu, Clara M.; Copot, Dana; De Keyser, Robin
• Is Part Of: IFAC-PapersOnLine. Volume 50:Issue 1(2017)
• Page Start: 15080
• Page End: 15085
• Publication Date: 2017-07
• Subjects: regulatory loops -- predictive control -- patient variability -- anaesthesia -- drug dosing control -- antagonist effect -- feedforward action -- disturbance rejection; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2017.08.2526
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: Many regulatory loops in drug delivery systems for depth of anesthesia optimization problem consider only the effect of the controller output on the patient pharmacokinetic and pharmacodynamic response. In reality, these drug assist devices are over-ruled by the anesthesiologist for setpoint changes, bolus intake and additional disturbances from the surgical team. Additionally, inter-patient variability imposes variations in the dynamic response and often intra-patient variability is also present. This paper introduces for the first time in literature a study on the effect of both controller and anesthesiologist in the loop for hypnosis regulation. Among the many control loops, model based predictive control is closest to mimic the anticipatory action of the anesthesiologist in real life and can actively deal with issues as time lags, delays, constraints, etc. This control algorithm is here combined with the action of the anesthesiologist. A disturbance signal to mimic surgical excitation has been introduced and a database of 25 patients has been derived from clinical insight. The results given in this paper reveal the antagonist effect in closed loop of the intervention from the anaesthesiologist when additional bolus intake is present. This observation explains induced dynamics in the closed loop observed in clinical trials and may be used as a starting point for next step in developing tools for improved assistance in clinical care.