An adaptive and generalizable closed-loop system for control of medically induced coma and other states of anesthesia

OBJECTIVE Design of closed-loop anesthetic delivery (CLAD) systems is an important topic, particularly for medically induced coma, which needs to be maintained for long periods. Current CLADs for medically induced coma require a separate offline experiment for model parameter estimation, which causes interruption in treatment and is difficult to perform. Also, CLADs may exhibit bias due to inherent time-variation and non-stationarity, and may have large infusion rate variations at steady state. Finally, current CLADs lack theoretical performance guarantees. We develop the first adaptive CLAD for medically induced coma, which addresses these limitations. Further, we extend our adaptive system to be generalizable to other states of anesthesia. APPROACH We designed general parametric pharmacodynamic, pharmacokinetic and neural observation models with associated guidelines, and derived a novel adaptive controller. We further penalized large steady-state drug infusion rate variations in the controller. We derived theoretical guarantees that the adaptive system has zero steady-state bias. Using simulations that resembled real time-varying and noisy environments, we tested the closed-loop system for control of two different anesthetic states, burst suppression in medically induced coma and unconsciousness in general anesthesia. MAIN RESULTS In 1200 simulations, the adaptive system achieved precise control of both anesthetic states despite non-stationarity, time-variation, noise, and no initial parameter knowledge. In both cases, the adaptive system performed close to a baseline system that knew the parameters exactly. In contrast, a non-adaptive system resulted in large steady-state bias and error. The adaptive system also resulted in significantly smaller steady-state infusion rate variations compared to prior systems. SIGNIFICANCE These results have significant implications for clinically viable CLAD design for a wide range of anesthetic states, with potential cost-saving and therapeutic benefits.

[1]  Steven L. Shafer,et al.  Algorithms to rapidly achieve and maintain stable drug concentrations at the site of drug effect with a computer-controlled infusion pump , 1992, Journal of Pharmacokinetics and Biopharmaceutics.

[2]  J. Schüttler,et al.  Model-Based Adaptive Control of Volatile Anesthetics by Quantitative EEG , 1995 .

[3]  S L Shafer,et al.  Pharmacokinetics and Pharmacodynamics of Propofol Infusions during General Anesthesia , 1988, Anesthesiology.

[4]  Dimitri P. Bertsekas,et al.  Dynamic Programming and Optimal Control, Two Volume Set , 1995 .

[5]  Morgan Le Guen,et al.  Feasibility of Closed-loop Titration of Propofol and Remifentanil Guided by the Spectral M-Entropy Monitor , 2012, Anesthesiology.

[6]  S L Shafer,et al.  Propofol Dosing Regimens for ICU Sedation Based upon an Integrated Pharmacokinetic– Pharmacodynamic Model , 2001, Anesthesiology.

[7]  Reginald G. Bickford,et al.  Automatic electroencephalographic control of general anesthesia , 1950 .

[8]  M H Eckman,et al.  Biologic Complexity in Mathematical Modeling , 1989, Medical decision making : an international journal of the Society for Medical Decision Making.

[9]  H Schwilden,et al.  Target controlled anaesthetic drug dosing. , 2008, Handbook of experimental pharmacology.

[10]  ShiNung Ching,et al.  Burst suppression probability algorithms: state-space methods for tracking EEG burst suppression , 2013, Journal of neural engineering.

[11]  E. Eger,et al.  Minimum alveolar anesthetic concentration: a standard of anesthetic potency. , 1965, Anesthesiology.

[12]  Jeffrey C. Sigl,et al.  The Effect of the Interaction of Propofol and Alfentanil on Recall, Loss of Consciousness, and the Bispectral Index , 1998 .

[13]  Emery N. Brown,et al.  Electroencephalogram signatures of loss and recovery of consciousness from propofol , 2013, Proceedings of the National Academy of Sciences.

[14]  A. Burm,et al.  Predictive performance of computer-controlled infusion of remifentanil during propofol/remifentanil anaesthesia. , 2003, British journal of anaesthesia.

[15]  Norbert Roewer,et al.  Is the bispectral index appropriate for monitoring the sedation level of mechanically ventilated surgical ICU patients? , 2002, Intensive Care Medicine.

[16]  H Schwilden,et al.  Closed‐loop Feedback Control of Methohexital Anesthesia by Quantitative EEG Analysis in Humans , 1987, Anesthesiology.

[17]  M. Struys,et al.  Comparison of Closed-loop Controlled Administration of Propofol Using Bispectral Index as the Controlled Variable versus “Standard Practice” Controlled Administration , 2001, Anesthesiology.

[18]  P. Sebel,et al.  Recovery of Consciousness after Thiopental or Propofol: Bispectral Index and the Isolated Forearm Technique , 1997, Anesthesiology.

[19]  G N Kenny,et al.  Closed-loop control of propofol anaesthesia. , 1999, British journal of anaesthesia.

[20]  Steven L. Shafer,et al.  Target controlled infusions: targeting the effect site while limiting peak plasma concentration , 2004, IEEE Transactions on Biomedical Engineering.

[21]  S. Shafer,et al.  The Influence of Method of Administration and Covariates on the Pharmacokinetics of Propofol in Adult Volunteers , 1998, Anesthesiology.

[22]  Paul E Dryden,et al.  Target-Controlled Infusions: Paths to Approval , 2016, Anesthesia and analgesia.

[23]  G N Kenny,et al.  Relationship between calculated blood concentration of propofol and electrophysiological variables during emergence from anaesthesia: comparison of bispectral index, spectral edge frequency, median frequency and auditory evoked potential index. , 1997, British journal of anaesthesia.

[24]  H. Schwilden,et al.  Prädiktivität und Präzision einer „target-controlled infusion” (TCI) von Propofol mit dem System „Disoprifusor TCI®” , 1998, Der Anaesthesist.

[25]  Joseph F. Cotten,et al.  Closed-loop Continuous Infusions of Etomidate and Etomidate Analogs in Rats: A Comparative Study of Dosing and the Impact on Adrenocortical Function , 2011, Anesthesiology.

[26]  Steven L Shafer,et al.  Target-Controlled Infusions: Surfing USA Redux. , 2016, Anesthesia and analgesia.

[27]  S L Shafer,et al.  Pharmacokinetics of Propofol in Adult Patients Undergoing Coronary Revascularization , 1996, Anesthesiology.

[28]  Petros A. Ioannou,et al.  Robust Adaptive Control , 2012 .

[29]  E. Brown,et al.  Clinical Electroencephalography for Anesthesiologists: Part I Background and Basic Signatures , 2015, Anesthesiology.

[30]  T. Hemmerling,et al.  Evaluation of a novel closed-loop total intravenous anaesthesia drug delivery system: a randomized controlled trial. , 2013, British journal of anaesthesia.

[31]  P. Steen,et al.  Bispectral index, predicted and measured drug levels of target‐controlled infusions of remifentanil and propofol during laparoscopic cholecystectomy and emergence , 2000, Acta anaesthesiologica Scandinavica.

[32]  J A Spain,et al.  Computer‐assisted Continuous Infusions of Fentanyl during Cardiac Anesthesia: Comparison with a Manual Method , 1985, Anesthesiology.

[33]  G Rolly,et al.  Closed‐loop controlled administration of propofol using bispectral analysis , 1998, Anaesthesia.

[34]  Ngai Liu,et al.  Closed-loop control of consciousness during lung transplantation: an observational study. , 2008, Journal of cardiothoracic and vascular anesthesia.

[35]  P J Manberg,et al.  Regulatory Challenges Facing Closed‐Loop Anesthetic Drug Infusion Devices , 2008, Clinical pharmacology and therapeutics.

[36]  R. Bickford,et al.  Electroencephalographically controlled anesthesia in abdominal surgery. , 1950, Journal of the American Medical Association.

[37]  E. Brown,et al.  General anesthesia, sleep, and coma. , 2010, The New England journal of medicine.

[38]  B. Matta,et al.  Burst suppression or isoelectric encephalogram for cerebral protection: evidence from metabolic suppression studies. , 1999, British journal of anaesthesia.

[39]  James C. Scott,et al.  Pharmacokinetics of fentanyl administered by computer-controlled infusion pump. , 1990, Anesthesiology.

[40]  Peter S. Sebel,et al.  Bispectral Analysis Measures Sedation and Memory Effects of Propofol, Midazolam, Isoflurane, and Alfentanil in Healthy Volunteers , 1997, Anesthesiology.

[41]  Samer Charabati,et al.  A randomized controlled trial demonstrates that a novel closed-loop propofol system performs better hypnosis control than manual administration , 2010, Canadian journal of anaesthesia = Journal canadien d'anesthesie.

[42]  M. Struys,et al.  Comparison of Plasma Compartment versus Two Methods for Effect Compartment–controlled Target-controlled Infusion for Propofol , 2000, Anesthesiology.

[43]  Julien Bogousslavsky,et al.  Propofol Treatment of Refractory Status Epilepticus: A Study of 31 Episodes , 2004, Epilepsia.

[44]  Laurent Beydon,et al.  Closed-Loop Coadministration of Propofol and Remifentanil Guided by Bispectral Index: A Randomized Multicenter Study , 2011, Anesthesia and analgesia.

[45]  I J Rampil,et al.  Anesthetic Potency (MAC) Is Independent of Forebrain Structures in the Rat , 1993, Anesthesiology.

[46]  D. Bogen,et al.  The Use of Computers for Controlling the Delivery of Anesthesia , 1992, Anesthesiology.

[47]  S L Shafer,et al.  Electromyographic activity falsely elevates the bispectral index. , 2000, Anesthesiology.

[48]  J.M. Bailey,et al.  A simple analytical solution to the three-compartment pharmacokinetic model suitable for computer-controlled infusion pumps , 1991, IEEE Transactions on Biomedical Engineering.

[49]  Wassim M. Haddad,et al.  Adaptive control for non-negative and compartmental dynamical systems with applications to general anesthesia , 2003 .

[50]  Emery N. Brown,et al.  A Brain-Machine Interface for Control of Medically-Induced Coma , 2013, PLoS Comput. Biol..

[51]  J. Glen,et al.  Pharmacokinetic model selection for target controlled infusions of propofol. Assessment of three parameter sets. , 1994 .

[52]  Ziv M. Williams,et al.  A Real-Time Brain-Machine Interface Combining Motor Target and Trajectory Intent Using an Optimal Feedback Control Design , 2013, PloS one.

[53]  D.A. Linkens,et al.  Adaptive and intelligent control in anesthesia , 1992, IEEE Control Systems.

[54]  Baruch Krauss,et al.  Can the bispectral index monitor quantify altered level of consciousness in emergency department patients? , 2003, Academic emergency medicine : official journal of the Society for Academic Emergency Medicine.

[55]  Q. C. Meng,et al.  A Conserved Behavioral State Barrier Impedes Transitions between Anesthetic-Induced Unconsciousness and Wakefulness: Evidence for Neural Inertia , 2010, PloS one.

[56]  L. Kearse,et al.  Bispectral Analysis of the Electroencephalogram Correlates with Patient Movement to Skin Incision during Propofol/Nitrous Oxide Anesthesia , 1994, Anesthesiology.

[57]  George A Mashour,et al.  Processed electroencephalogram in depth of anesthesia monitoring , 2009, Current opinion in anaesthesiology.

[58]  G D Puri,et al.  Closed-Loop Anaesthesia Delivery System (CLADSTM) using Bispectral Index: A Performance Assessment Study , 2007, Anaesthesia and intensive care.

[59]  D. Chernik,et al.  Validity and Reliability of the Observer's: Assessment of Alertness/Sedation Scale Study with Intravenous Midazolam , 1990, Journal of clinical psychopharmacology.

[60]  P C Vijn,et al.  I.v. anaesthesia and EEG burst suppression in rats: bolus injections and closed-loop infusions. , 1998, British journal of anaesthesia.

[61]  J. Agarwal,et al.  Comparison of closed loop vs. manual administration of propofol using the Bispectral index in cardiac surgery , 2009, Acta anaesthesiologica Scandinavica.

[62]  Luc Barvais,et al.  Titration of Propofol for Anesthetic Induction and Maintenance Guided by the Bispectral Index: Closed-loop versus Manual Control: A Prospective, Randomized, Multicenter Study , 2006, Anesthesiology.

[63]  J G Reves,et al.  Pharmacokinetic model-driven infusion of fentanyl: assessment of accuracy. , 1990, Anesthesiology.

[64]  S L Shafer,et al.  A comparison of spectral edge, delta power, and bispectral index as EEG measures of alfentanil, propofol, and midazolam drug effect , 1997, Clinical pharmacology and therapeutics.

[65]  H. Schwilden,et al.  A general method for calculating the dosage scheme in linear pharmacokinetics , 2004, European Journal of Clinical Pharmacology.

[66]  J. Antognini,et al.  Exaggerated Anesthetic Requirements in the Preferentially Anesthetized Brain , 1993, Anesthesiology.

[67]  E. Brown,et al.  General anesthesia and altered states of arousal: a systems neuroscience analysis. , 2011, Annual review of neuroscience.

[68]  Peter S A Glass.  Automated control of anesthesia ten years later: futuristic novelty or present day reality , 2010, Canadian journal of anaesthesia = Journal canadien d'anesthesie.

[69]  Maxime Cannesson,et al.  Review article: closed-loop systems in anesthesia: is there a potential for closed-loop fluid management and hemodynamic optimization? , 2012, Anesthesia and analgesia.

[70]  William J Jusko,et al.  Diversity of mechanism-based pharmacodynamic models. , 2003, Drug metabolism and disposition: the biological fate of chemicals.

[71]  Emery N. Brown,et al.  Neural population partitioning and a concurrent brain-machine interface for sequential motor function , 2012, Nature Neuroscience.

[72]  K. Leslie,et al.  Target-Controlled Infusion Versus Manually-Controlled Infusion of Propofol for General Anaesthesia or Sedation in Adults , 2008, Anesthesia and analgesia.

[73]  H. Schwilden,et al.  Closed-loop feedback control of propofol anaesthesia by quantitative EEG analysis in humans. , 1989, British journal of anaesthesia.

[74]  Petre Stoica,et al.  Decentralized Control , 2018, The Control Systems Handbook.

[75]  K. Han,et al.  Bispectral Index Does Not Correlate with Observer Assessment of Alertness and Sedation Scores During 0.5% Bupivacaine Epidural Anesthesia with Nitrous Oxide Sedation , 2006, Anesthesia and analgesia.

[76]  Emery N Brown,et al.  A neurophysiological–metabolic model for burst suppression , 2012, Proceedings of the National Academy of Sciences.

[77]  Michel M. R. F. Struys,et al.  Estimation of Optimal Modeling Weights for a Bayesian-Based Closed-Loop System for Propofol Administration Using the Bispectral Index as a Controlled Variable: A Simulation Study , 2007, Anesthesia and analgesia.

[78]  A. León,et al.  Modelling propofol pharmacodynamics using BIS‐guided anaesthesia , 2013, Anaesthesia.

[79]  I. Kissin,et al.  Depth of Anesthesia and Bispectral Index Monitoring , 2000, Anesthesia and analgesia.

[80]  Naira Hovakimyan,et al.  Neural Network Adaptive Output Feedback Control for Intensive Care Unit Sedation and Intraoperative Anesthesia , 2007, IEEE Transactions on Neural Networks.

[81]  Emery N Brown,et al.  A closed-loop anesthetic delivery system for real-time control of burst suppression , 2013, Journal of neural engineering.

[82]  D. Pfaff,et al.  Recovery of consciousness is mediated by a network of discrete metastable activity states , 2014, Proceedings of the National Academy of Sciences.

[83]  ShiNung Ching,et al.  Robust control of burst suppression for medical coma , 2015, Journal of neural engineering.

[84]  K. Ikeda,et al.  Electroencephalographic Derivatives as a Tool for Predicting the Depth of Sedation and Anesthesia Induced by Sevoflurane , 1998, Anesthesiology.

[85]  Michel M. R. F. Struys,et al.  The Accuracy and Clinical Feasibility of a New Bayesian-Based Closed-Loop Control System for Propofol Administration Using the Bispectral Index as a Controlled Variable , 2008, Anesthesia and analgesia.

[86]  S. Nasraway,et al.  How reliable is the Bispectral Index in critically ill patients? A prospective, comparative, single-blinded observer study* , 2002, Critical care medicine.

[87]  Rob J. Roy,et al.  Adaptive control of closed-circuit anesthesia , 1991 .

[88]  Steven L Shafer,et al.  Target-controlled Infusions for Intravenous Anesthetics Surfing USA Not ! , 2003 .

[89]  Peter S. Sebel,et al.  Eeg bispectrum predicts movement during thiopental/isoflurane anesthesia , 1995, Journal of Clinical Monitoring.

[90]  G. W. Wornell,et al.  Feedback-Controlled Parallel Point Process Filter for Estimation of Goal-Directed Movements From Neural Signals , 2013, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[91]  Florin Amzica,et al.  Basic physiology of burst‐suppression , 2009, Epilepsia.

[92]  Steven L. Shafer,et al.  Measuring the predictive performance of computer-controlled infusion pumps , 1992, Journal of Pharmacokinetics and Biopharmaceutics.

[93]  ShiNung Ching,et al.  Real-time Closed-loop Control in a Rodent Model of Medically Induced Coma Using Burst Suppression , 2013, Anesthesiology.