Accurate detection of spontaneous seizures using a generalized linear model with external validation.

OBJECTIVE: Seizure detection is a major facet of electroencephalography (EEG) analysis in neurocritical care, epilepsy diagnosis and management, and the instantiation of novel therapies such as closed-loop stimulation or optogenetic control of seizures. It is also of increased importance in high-throughput, robust, and reproducible pre-clinical research. However, seizure detectors are not widely relied upon in either clinical or research settings due to limited validation. In this study, we create a high-performance seizure-detection approach, validated in multiple data sets, with the intention that such a system could be available to users for multiple purposes. METHODS: We introduce a generalized linear model trained on 141 EEG signal features for classification of seizures in continuous EEG for two data sets. In the first (Focal Epilepsy) data set consisting of 16 rats with focal epilepsy, we collected 1012 spontaneous seizures over 3 months of 24/7 recording. We trained a generalized linear model on the 141 features representing 20 feature classes, including univariate and multivariate, linear and nonlinear, time, and frequency domains. We tested performance on multiple hold-out test data sets. We then used the trained model in a second (Multifocal Epilepsy) data set consisting of 96 rats with 2883 spontaneous multifocal seizures. RESULTS: From the Focal Epilepsy data set, we built a pooled classifier with an Area Under the Receiver Operating Characteristic (AUROC) of 0.995 and leave-one-out classifiers with an AUROC of 0.962. We validated our method within the independently constructed Multifocal Epilepsy data set, resulting in a pooled AUROC of 0.963. We separately validated a model trained exclusively on the Focal Epilepsy data set and tested on the held-out Multifocal Epilepsy data set with an AUROC of 0.890. Latency to detection was under 5 seconds for over 80% of seizures and under 12 seconds for over 99% of seizures. SIGNIFICANCE: This method achieves the highest performance published for seizure detection on multiple independent data sets. This method of seizure detection can be applied to automated EEG analysis pipelines as well as closed loop interventional approaches, and can be especially useful in the setting of research using animals in which there is an increased need for standardization and high-throughput analysis of large number of seizures.

A metabolomics-based molecular pathway analysis of how the sodium-glucose co-transporter-2 inhibitor dapagliflozin may slow kidney function decline in patients with diabetes.

AIM: To investigate which metabolic pathways are targeted by the sodium-glucose co-transporter-2 inhibitor dapagliflozin to explore the molecular processes involved in its renal protective effects. METHODS: An unbiased mass spectrometry plasma metabolomics assay was performed on baseline and follow-up (week 12) samples from the EFFECT II trial in patients with type 2 diabetes with non-alcoholic fatty liver disease receiving dapagliflozin 10 mg/day (n = 19) or placebo (n = 6). Transcriptomic signatures from tubular compartments were identified from kidney biopsies collected from patients with diabetic kidney disease (DKD) (n = 17) and healthy controls (n = 30) from the European Renal cDNA Biobank. Serum metabolites that significantly changed after 12 weeks of dapagliflozin were mapped to a metabolite-protein interaction network. These proteins were then linked with intra-renal transcripts that were associated with DKD or estimated glomerular filtration rate (eGFR). The impacted metabolites and their protein-coding transcripts were analysed for enriched pathways. RESULTS: Of all measured (n = 812) metabolites, 108 changed (P < 0.05) during dapagliflozin treatment and 74 could be linked to 367 unique proteins/genes. Intra-renal mRNA expression analysis of the genes encoding the metabolite-associated proteins using kidney biopsies resulted in 105 genes that were significantly associated with eGFR in patients with DKD, and 135 genes that were differentially expressed between patients with DKD and controls. The combination of metabolites and transcripts identified four enriched pathways that were affected by dapagliflozin and associated with eGFR: glycine degradation (mitochondrial function), TCA cycle II (energy metabolism), L-carnitine biosynthesis (energy metabolism) and superpathway of citrulline metabolism (nitric oxide synthase and endothelial function). CONCLUSION: The observed molecular pathways targeted by dapagliflozin and associated with DKD suggest that modifying molecular processes related to energy metabolism, mitochondrial function and endothelial function may contribute to its renal protective effect.

Predicting short- and long-term glycated haemoglobin response after insulin initiation in patients with type 2 diabetes mellitus using machine-learning algorithms.

AIM: To assess the potential of supervised machine-learning techniques to identify clinical variables for predicting short-term and long-term glycated haemoglobin (HbA1c) response after insulin treatment initiation in patients with type 2 diabetes mellitus (T2DM). MATERIALS AND METHODS: We included patients with T2DM from the Groningen Initiative to Analyse Type 2 diabetes Treatment (GIANTT) database who started insulin treatment between 2007 and 2013 and had a minimum follow-up of 2 years. Short- and long-term responses at 6 (±2) and 24 (±2) months after insulin initiation, respectively, were assessed. Patients were defined as good responders if they had a decrease in HbA1c ≥ 5 mmol/mol or reached the recommended level of HbA1c ≤ 53 mmol/mol. Twenty-four baseline clinical variables were used for the analysis and an elastic net regularization technique was used for variable selection. The performance of three traditional machine-learning algorithms was compared for the prediction of short- and long-term responses and the area under the receiver-operating characteristic curve (AUC) was used to assess the performance of the prediction models. RESULTS: The elastic net regularization-based generalized linear model, which included baseline HbA1c and estimated glomerular filtration rate, correctly classified short- and long-term HbA1c response after treatment initiation, with AUCs of 0.80 (95% CI 0.78-0.83) and 0.81 (95% CI 0.79-0.84), respectively, and outperformed the other machine-learning algorithms. Using baseline HbA1c alone, an AUC = 0.71 (95% CI 0.65-0.73) and 0.72 (95% CI 0.66-0.75) was obtained for predicting short-term and long-term response, respectively. CONCLUSIONS: Machine-learning algorithm performed well in the prediction of an individual's short-term and long-term HbA1c response using baseline clinical variables.

Novel drug-independent sedation level estimation based on machine learning of quantitative frontal electroencephalogram features in healthy volunteers.

BACKGROUND: Sedation indicators based on a single quantitative EEG (QEEG) feature have been criticised for their limited performance. We hypothesised that integration of multiple QEEG features into a single sedation-level estimator using a machine learning algorithm could reliably predict levels of sedation, independent of the sedative drug used. METHODS: In total, 102 subjects receiving propofol (N=36; 16 male/20 female), sevoflurane (N=36; 16 male/20 female), or dexmedetomidine (N=30; 15 male/15 female) were included in this study of healthy volunteers. Sedation level was assessed using the Modified Observer's Assessment of Alertness/Sedation (MOAA/S) score. We used 44 QEEG features estimated from the EEG data in a logistic regression algorithm, and an elastic-net regularisation method was used for feature selection. The area under the receiver operator characteristic curve (AUC) was used to assess the performance of the logistic regression model. RESULTS: The performances obtained when the system was trained and tested as drug-dependent mode to distinguish between awake and sedated states (mean AUC [standard deviation]) were propofol=0.97 (0.03), sevoflurane=0.74 (0.25), and dexmedetomidine=0.77 (0.10). The drug-independent system resulted in mean AUC=0.83 (0.17) to discriminate between the awake and sedated states. CONCLUSIONS: The incorporation of large numbers of QEEG features and machine learning algorithms is feasible for next-generation monitors of sedation level. Different QEEG features were selected for propofol, sevoflurane, and dexmedetomidine groups, but the sedation-level estimator maintained a high performance for predicting MOAA/S independent of the drug used. CLINICAL TRIAL REGISTRATION: NCT02043938; NCT03143972.

The Neu-Prem Trial: Neuromonitoring of Brains of Infants Born Preterm During Resuscitation-A Prospective Observational Cohort Study.

OBJECTIVE: To determine whether monitoring cerebral oxygen tissue saturation (StO2) with near-infrared spectroscopy (NIRS) and brain activity with amplitude-integrated electroencephalography (aEEG) can predict infants at risk for intraventricular hemorrhage (IVH) and death in the first 72 hours of life. STUDY DESIGN: A NIRS sensor and electroencephalography leads were placed on 127 newborns <32 weeks of gestational age at birth. Ten minutes of continuous NIRS and aEEG along with heart rate, peripheral arterial oxygen saturation, fraction of inspired oxygen, and mean airway pressure measurements were obtained in the delivery room. Once the infant was transferred to the neonatal intensive care unit, NIRS, aEEG, and vital signs were recorded until 72 hours of life. An ultrasound scan of the head was performed within the first 12 hours of life and again at 72 hours of life. RESULTS: Thirteen of the infants developed any IVH or died; of these, 4 developed severe IVH (grade 3-4) within 72 hours. There were no differences in either cerebral StO2 or aEEG in the infants with low-grade IVH. Infants who developed severe IVH or death had significantly lower cerebral StO2 from 8 to 10 minutes of life. CONCLUSIONS: aEEG was not predictive of IVH or death in the delivery room or in the neonatal intensive care unit. It may be possible to use NIRS in the delivery room to predict severe IVH and early death. TRIAL REGISTRATION: ClinicalTrials.gov: NCT02605733.

ADARRI: a novel method to detect spurious R-peaks in the electrocardiogram for heart rate variability analysis in the intensive care unit.

We developed a simple and fully automated method for detecting artifacts in the R-R interval (RRI) time series of the ECG that is tailored to the intensive care unit (ICU) setting. From ECG recordings of 50 adult ICU-subjects we selected 60 epochs with valid R-peak detections and 60 epochs containing artifacts leading to missed or false positive R-peak detections. Next, we calculated the absolute value of the difference between two adjacent RRIs (adRRI), and obtained the empirical probability distributions of adRRI values for valid R-peaks and artifacts. From these, we calculated an optimal threshold for separating adRRI values arising from artifact versus non-artefactual data. We compared the performance of our method with the methods of Berntson and Clifford on the same data. We identified 257,458 R-peak detections, of which 235,644 (91.5%) were true detections and 21,814 (8.5%) arose from artifacts. Our method showed superior performance for detecting artifacts with sensitivity 100%, specificity 99%, precision 99%, positive likelihood ratio of 100 and negative likelihood ratio <0.001 compared to Berntson's and Clifford's method with a sensitivity, specificity, precision and positive and negative likelihood ratio of 99%, 78%, 82%, 4.5, 0.013 for Berntson's method and 55%, 98%, 96%, 27.5, 0.460 for Clifford's method, respectively. A novel algorithm using a patient-independent threshold derived from the distribution of adRRI values in ICU ECG data identifies artifacts accurately, and outperforms two other methods in common use. Furthermore, the threshold was calculated based on real data from critically ill patients and the algorithm is easy to implement.

Patient-Specific Classification of ICU Sedation Levels From Heart Rate Variability.

OBJECTIVE: To develop a personalizable algorithm to discriminate between sedation levels in ICU patients based on heart rate variability. DESIGN: Multicenter, pilot study. SETTING: Several ICUs at Massachusetts General Hospital, Boston, MA. PATIENTS: We gathered 21,912 hours of routine electrocardiogram recordings from a heterogenous group of 70 adult ICU patients. All patients included in the study were mechanically ventilated and were receiving sedatives. MEASUREMENTS AND MAIN RESULTS: As "ground truth" for developing our method, we used Richmond Agitation Sedation Scale scores grouped into four levels denoted "comatose" (-5), "deeply sedated" (-4 to -3), "lightly sedated" (-2 to 0), and "agitated" (+1 to +4). We trained a support vector machine learning algorithm to calculate the probability of each sedation level from heart rate variability measures derived from the electrocardiogram. To estimate algorithm performance, we calculated leave-one-subject out cross-validated accuracy. The patient-independent version of the proposed system discriminated between the four sedation levels with an overall accuracy of 59%. Upon personalizing the system supplementing the training data with patient-specific calibration data, consisting of an individual's labeled heart rate variability epochs from the preceding 24 hours, accuracy improved to 67%. The personalized system discriminated between light- and deep-sedation states with an average accuracy of 75%. CONCLUSIONS: With further refinement, the methodology reported herein could lead to a fully automated system for depth of sedation monitoring. By enabling monitoring to be continuous, such technology may help clinical staff to monitor sedation levels more effectively and to reduce complications related to over- and under sedation.

Automatic Classification of Sedation Levels in ICU Patients Using Heart Rate Variability.

OBJECTIVE: To explore the potential value of heart rate variability features for automated monitoring of sedation levels in mechanically ventilated ICU patients. DESIGN: Multicenter, pilot study. SETTING: Several ICUs at Massachusetts General Hospital, Boston, MA. PATIENTS: Electrocardiogram recordings from 40 mechanically ventilated adult patients receiving sedatives in an ICU setting were used to develop and test the proposed automated system. MEASUREMENTS AND MAIN RESULTS: Richmond Agitation-Sedation Scale scores were acquired prospectively to assess patient sedation levels and were used as ground truth. Richmond Agitation-Sedation Scale scores were grouped into four levels, denoted "unarousable" (Richmond Agitation- Sedation Scale = -5, -4), "sedated" (-3, -2, -1), "awake" (0), "agitated" (+1, +2, +3, +4). A multiclass support vector machine algorithm was used for classification. Classifier training and performance evaluations were carried out using leave-oneout cross validation. An overall accuracy of 69% was achieved for discriminating between the four levels of sedation. The proposed system was able to reliably discriminate (accuracy = 79%) between sedated (Richmond Agitation-Sedation Scale < 0) and nonsedated states (Richmond Agitation-Sedation Scale > 0). CONCLUSIONS: With further refinement, the methodology reported herein could lead to a fully automated system for depth of sedation monitoring. By enabling monitoring to be continuous, such technology may help clinical staff to monitor sedation levels more effectively and to reduce complications related to over- and undersedation.

Do all sedatives promote biological sleep electroencephalogram patterns? A machine learning framework to identify biological sleep promoting sedatives using electroencephalogram.

BACKGROUND: Sedatives are commonly used to promote sleep in intensive care unit patients. However, it is not clear whether sedation-induced states are similar to the biological sleep. We explored if sedative-induced states resemble biological sleep using multichannel electroencephalogram (EEG) recordings. METHODS: Multichannel EEG datasets from two different sources were used in this study: (1) sedation dataset consisting of 102 healthy volunteers receiving propofol (N = 36), sevoflurane (N = 36), or dexmedetomidine (N = 30), and (2) publicly available sleep EEG dataset (N = 994). Forty-four quantitative time, frequency and entropy features were extracted from EEG recordings and were used to train the machine learning algorithms on sleep dataset to predict sleep stages in the sedation dataset. The predicted sleep states were then compared with the Modified Observer's Assessment of Alertness/ Sedation (MOAA/S) scores. RESULTS: The performance of the model was poor (AUC = 0.55-0.58) in differentiating sleep stages during propofol and sevoflurane sedation. In the case of dexmedetomidine, the AUC of the model increased in a sedation-dependent manner with NREM stages 2 and 3 highly correlating with deep sedation state reaching an AUC of 0.80. CONCLUSIONS: We addressed an important clinical question to identify biological sleep promoting sedatives using EEG signals. We demonstrate that propofol and sevoflurane do not promote EEG patterns resembling natural sleep while dexmedetomidine promotes states resembling NREM stages 2 and 3 sleep, based on current sleep staging standards.

Dexmedetomidine-induced deep sedation mimics non-rapid eye movement stage 3 sleep: large-scale validation using machine learning.

STUDY OBJECTIVES: Dexmedetomidine-induced electroencephalogram (EEG) patterns during deep sedation are comparable with natural sleep patterns. Using large-scale EEG recordings and machine learning techniques, we investigated whether dexmedetomidine-induced deep sedation indeed mimics natural sleep patterns. METHODS: We used EEG recordings from three sources in this study: 8,707 overnight sleep EEG and 30 dexmedetomidine clinical trial EEG. Dexmedetomidine-induced sedation levels were assessed using the Modified Observer's Assessment of Alertness/Sedation (MOAA/S) score. We extracted 22 spectral features from each EEG recording using a multitaper spectral estimation method. Elastic-net regularization method was used for feature selection. We compared the performance of several machine learning algorithms (logistic regression, support vector machine, and random forest), trained on individual sleep stages, to predict different levels of the MOAA/S sedation state. RESULTS: The random forest algorithm trained on non-rapid eye movement stage 3 (N3) predicted dexmedetomidine-induced deep sedation (MOAA/S = 0) with area under the receiver operator characteristics curve >0.8 outperforming other machine learning models. Power in the delta band (0-4 Hz) was selected as an important feature for prediction in addition to power in theta (4-8 Hz) and beta (16-30 Hz) bands. CONCLUSIONS: Using a large-scale EEG data-driven approach and machine learning framework, we show that dexmedetomidine-induced deep sedation state mimics N3 sleep EEG patterns. CLINICAL TRIALS: Name-Pharmacodynamic Interaction of REMI and DMED (PIRAD), URL-https://clinicaltrials.gov/ct2/show/NCT03143972, and registration-NCT03143972.

Automated tracking of level of consciousness and delirium in critical illness using deep learning.

Over- and under-sedation are common in the ICU, and contribute to poor ICU outcomes including delirium. Behavioral assessments, such as Richmond Agitation-Sedation Scale (RASS) for monitoring levels of sedation and Confusion Assessment Method for the ICU (CAM-ICU) for detecting signs of delirium, are often used. As an alternative, brain monitoring with electroencephalography (EEG) has been proposed in the operating room, but is challenging to implement in ICU due to the differences between critical illness and elective surgery, as well as the duration of sedation. Here we present a deep learning model based on a combination of convolutional and recurrent neural networks that automatically tracks both the level of consciousness and delirium using frontal EEG signals in the ICU. For level of consciousness, the system achieves a median accuracy of 70% when allowing prediction to be within one RASS level difference across all patients, which is comparable or higher than the median technician-nurse agreement at 59%. For delirium, the system achieves an AUC of 0.80 with 69% sensitivity and 83% specificity at the optimal operating point. The results show it is feasible to continuously track level of consciousness and delirium in the ICU.

Quantitative EEG reactivity and machine learning for prognostication in hypoxic-ischemic brain injury.

OBJECTIVE: Electroencephalogram (EEG) reactivity is a robust predictor of neurological recovery after cardiac arrest, however interrater-agreement among electroencephalographers is limited. We sought to evaluate the performance of machine learning methods using EEG reactivity data to predict good long-term outcomes in hypoxic-ischemic brain injury. METHODS: We retrospectively reviewed clinical and EEG data of comatose cardiac arrest subjects. Electroencephalogram reactivity was tested within 72 h from cardiac arrest using sound and pain stimuli. A Quantitative EEG (QEEG) reactivity method evaluated changes in QEEG features (EEG spectra, entropy, and frequency features) during the 10 s before and after each stimulation. Good outcome was defined as Cerebral Performance Category of 1-2 at six months. Performance of a random forest classifier was compared against a penalized general linear model (GLM) and expert electroencephalographer review. RESULTS: Fifty subjects were included and sixteen (32%) had good outcome. Both QEEG reactivity methods had comparable performance to expert EEG reactivity assessment for good outcome prediction (mean AUC 0.8 for random forest vs. 0.69 for GLM vs. 0.69 for expert review, respectively; p non-significant). CONCLUSIONS: Machine-learning models utilizing quantitative EEG reactivity data can predict long-term outcome after cardiac arrest. SIGNIFICANCE: A quantitative approach to EEG reactivity assessment may support prognostication in cardiac arrest.

Automatic Detection of General Anesthetic-States using ECG-Derived Autonomic Nervous System Features.

Electroencephalogram (EEG)-based prediction systems are used to target anesthetic-states in patients undergoing procedures with general anesthesia (GA). These systems are not widely employed in resource-limited settings because they are cost-prohibitive. Although anesthetic-drugs induce highly-structured, oscillatory neural dynamics that make EEG-based systems a principled approach for anesthetic-state monitoring, anesthetic-drugs also significantly modulate the autonomic nervous system (ANS). Because ANS dynamics can be inferred from electrocardiogram (ECG) features such as heart rate variability, it may be possible to develop an ECG-based system to infer anesthetic-states as a low-cost and practical alternative to EEG-based anesthetic-state prediction systems. In this work, we demonstrate that an ECG-based system using ANS features can be used to discriminate between non-GA and GA states in sevoflurane, with a GA F1 score of 0.834, [95% CI, 0.776, 0.892], and in sevoflurane-plus-ketamine, with a GA F1 score of 0.880 [0.815, 0.954]. With further refinement, ECG-based anesthetic-state systems could be developed as a fully automated system for anesthetic-state monitoring in resource-limited settings.

Brain Monitoring of Sedation in the Intensive Care Unit Using a Recurrent Neural Network.

Over and under-sedation are common in critically ill patients admitted to the Intensive Care Unit. Clinical assessments provide limited time resolution and are based on behavior rather than the brain itself. Existing brain monitors have been developed primarily for non-ICU settings. Here, we use a clinical dataset from 154 ICU patients in whom the Richmond Agitation-Sedation Score is assessed about every 2 hours. We develop a recurrent neural network (RNN) model to discriminate between deep vs. no sedation, trained end-to-end from raw EEG spectrograms without any feature extraction. We obtain an average area under the ROC of 0.8 on 10-fold cross validation across patients. Our RNN is able to provide reliable estimates of sedation levels consistently better compared to a feed-forward model with simple smoothing. Decomposing the prediction error in terms of sedatives reveals that patient-specific calibration for sedatives is expected to further improve sedation monitoring.

The revised Cerebral Recovery Index improves predictions of neurological outcome after cardiac arrest.

OBJECTIVE: Analysis of the electroencephalogram (EEG) background pattern helps predicting neurological outcome of comatose patients after cardiac arrest (CA). Visual analysis may not extract all discriminative information. We present predictive values of the revised Cerebral Recovery Index (rCRI), based on continuous extraction and combination of a large set of evolving quantitative EEG (qEEG) features and machine learning techniques. METHODS: We included 551 subsequent patients from a prospective cohort study on continuous EEG after CA in two hospitals. Outcome at six months was classified as good (Cerebral Performance Category (CPC) 1-2) or poor (CPC 3-5). Forty-four qEEG features (from time, frequency and entropy domain) were selected by the least absolute shrinkage and selection operator (LASSO) method and used in a Random Forests classification system. We trained and evaluated the system with 10-fold cross validation. For poor outcome prediction, the sensitivity at 100% specificity (Se100) and the area under the receiver operator curve (AUC) were used as performance of the prediction model. For good outcome, we used the sensitivity at 95% specificity (Se95). RESULTS: Two hundred fifty-six (47%) patients had a good outcome. The rCRI predicted poor outcome with AUC = 0.94 (95% CI: 0.83-0.91), Se100 = 0.66 (0.65-0.78), and AUC = 0.88 (0.78-0.93), Se100 = 0.60 (0.51-0.75) at 12 and 24 h after CA, respectively. The rCRI predicted good outcome with Se95 = 0.72 (0.61-0.85) and 0.40 (0.30-0.51) at 12 and 24 h after CA, respectively. CONCLUSIONS: Results obtained in this study suggest that with machine learning algorithms and large set of qEEG features, it is possible to efficiently monitor patient outcome after CA. We also demonstrate the importance of selection of optimal performance metric to train a classifier model for outcome prediction. SIGNIFICANCE: The rCRI is a sensitive, reliable predictor of neurological outcome of comatose patients after CA.

Heart rate variability as a biomarker for sedation depth estimation in ICU patients.

An automated patient-specific system to classify the level of sedation in ICU patients using heart rate variability signal is presented in this paper. ECG from 70 mechanically ventilated adult patients with administered sedatives in an ICU setting were used to develop a support vector machine based system for sedation depth monitoring using several heart rate variability measures. A leave-one-subject-out cross validation was used for classifier training and performance evaluations. The proposed patient-specific system provided a sensitivity, specificity and an AUC of 64%, 84.8% and 0.72, respectively. It is hoped that with the help of additional physiological signals the proposed patient-specific sedation level prediction system could lead to a fully automated multimodal system to assist clinical staff in ICUs to interpret the sedation level of the patient.

Clustering analysis to identify distinct spectral components of encephalogram burst suppression in critically ill patients.

Millions of patients are admitted each year to intensive care units (ICUs) in the United States. A significant fraction of ICU survivors develop life-long cognitive impairment, incurring tremendous financial and societal costs. Delirium, a state of impaired awareness, attention and cognition that frequently develops during ICU care, is a major risk factor for post-ICU cognitive impairment. Recent studies suggest that patients experiencing electroencephalogram (EEG) burst suppression have higher rates of mortality and are more likely to develop delirium than patients who do not experience burst suppression. Burst suppression is typically associated with coma and deep levels of anesthesia or hypothermia, and is defined clinically as an alternating pattern of high-amplitude "burst" periods interrupted by sustained low-amplitude "suppression" periods. Here we describe a clustering method to analyze EEG spectra during burst and suppression periods. We used this method to identify a set of distinct spectral patterns in the EEG during burst and suppression periods in critically ill patients. These patterns correlate with level of patient sedation, quantified in terms of sedative infusion rates and clinical sedation scores. This analysis suggests that EEG burst suppression in critically ill patients may not be a single state, but instead may reflect a plurality of states whose specific dynamics relate to a patient's underlying brain function.