Uncommon EEG burst-suppression in severe postanoxic encephalopathy

ObjectiveIn patients suffering from severe hypoxia, the EEG may show a burst-suppression pattern, characterized by low-voltage activity and the occurrence of high amplitude burst-like events. We describe the two-timescale burst phenomenology of this postanoxic condition.MethodsWe present EEG recordings showing remarkable burst phenomenology in two postanoxic patients and consider potential mechanisms responsible for the generation of the burst-suppression patterns. We quantify the postanoxic condition in terms of the dimension (number of degrees of freedom) of its dynamics by comparing our data with a system of three ordinary differential equations with two timescales subject to varying degrees of noise.ResultsEEGs displayed extreme similarity of the bursts, separated by interburst intervals up to more than 300 s. This pattern reflects a significant reduction in the number of functional brain states. This post-anoxic condition is found to have dimension 3, consisting of fast dynamics responsible for the bifurcation to bursting behavior, and a long time-scale responsible for burst termination and the interburst intervals.ConclusionsLow-dimensional postanoxic brain states, as manifested by burst-similarity, appears to indicate an irreversible loss of brain function and consciousness.SignificanceEvidence of brain functionality in a persistent low dimensional state due to severe hypoxia is indicative of permanent loss of consciousness with essentially no chance for recovery. Quantitative evidence for such degenerate states is important for clinical decision making.     

The prognostic value of discontinuous EEG patterns in postanoxic coma

Objective

To assess the value of background continuity and amplitude fluctuations of the EEG for the prediction of outcome of comatose patients after cardiac arrest.

Periodic eye opening and swallowing movements associated with post-anoxic burst-suppression EEG pattern.

We report a video-electroencephalogram study of a 50-year-old woman who developed a clinical picture consisting of stereotyped periodic eye opening followed by eye closing, with or without swallowing movements, after a prolonged cardiopulmonary arrest. These movements were associated with a burst-suppression pattern on the electroencephalogram. [Published with video sequences].     

EEG functional connectivity contributes to outcome prediction of postanoxic coma

ObjectiveTo investigate the additional value of EEG functional connectivity features, in addition to non-coupling EEG features, for outcome prediction of comatose patients after cardiac arrest.MethodsProspective, multicenter cohort study. Coherence, phase locking value, and mutual information were calculated in 19-channel EEGs at 12 h, 24 h and 48 h after cardiac arrest. Three sets of machine learning classification models were trained and validated with functional connectivity, EEG non-coupling features, and a combination of these. Neurological outcome was assessed at six months and categorized as “good” (Cerebral Performance Category [CPC] 1–2) or “poor” (CPC 3–5).ResultsWe included 594 patients (46% good outcome). A sensitivity of 51% (95% CI: 34–56%) at 100% specificity in predicting poor outcome was achieved by the best functional connectivity-based classifier at 12 h after cardiac arrest, while the best non-coupling-based model reached a sensitivity of 32% (0–54%) at 100% specificity using data at 12 h and 48 h. Combination of both sets of features achieved a sensitivity of 73% (50–77%) at 100% specificity.ConclusionFunctional connectivity measures improve EEG based prediction models for poor outcome of postanoxic coma.SignificanceFunctional connectivity features derived from early EEG hold potential to improve outcome prediction of coma after cardiac arrest.     

Dynamic functional connectivity of the EEG in relation to outcome of postanoxic coma

ObjectiveEarly EEG contains reliable information for outcome prediction of comatose patients after cardiac arrest. We introduce dynamic functional connectivity measures and estimate additional predictive values.MethodsWe performed a prospective multicenter cohort study on continuous EEG for outcome prediction of comatose patients after cardiac arrest. We calculated Link Rates (LR) and Link Durations (LD) in the α, δ, and θ band, based on similarity of instantaneous frequencies in five-minute EEG epochs, hourly, during 3 days after cardiac arrest. We studied associations of LR and LD with good (Cerebral Performance Category (CPC) 1–2) or poor outcome (CPC 3–5) with univariate analyses. With random forest classification, we established EEG-based predictive models. We used receiver operating characteristics to estimate additional values of dynamic connectivity measures for outcome prediction.ResultsOf 683 patients, 369 (54%) had poor outcome. Patients with poor outcome had significantly lower LR and longer LD, with largest differences 12 h after cardiac arrest (LR_θ 1.87 vs. 1.95 Hz and LD_α 91 vs. 82 ms). Adding these measures to a model with classical EEG features increased sensitivity for reliable prediction of poor outcome from 34% to 38% at 12 h after cardiac arrest.ConclusionPoor outcome is associated with lower dynamics of connectivity after cardiac arrest.SignificanceDynamic functional connectivity analysis may improve EEG based outcome prediction.     

Intracranial EEG in temporal lobe epilepsy.

Intracranial EEG monitoring before epilepsy surgery, while becoming less commonly performed in patients with unilateral mesial temporal lobe epilepsy, is still widely used when bilateral independent temporal lobe seizures are suspected or when extratemporal foci cannot be ruled out by noninvasive means. Additionally, many epilepsy centers are reporting excellent surgical outcome in patients with neocortical temporal lobe epilepsy, when resections are guided by intracranial EEG studies. This article reviews the indications, technical aspects, risks, and interpretation of intracranial EEG in patients with temporal lobe seizures. It also considers intracranial EEG features predictive of surgical outcome.     

Detection of EEG burst-suppression in neurocritical care patients using an unsupervised machine learning algorithm

ObjectiveThe burst suppression pattern in clinical electroencephalographic (EEG) recordings is an important diagnostic tool because of its association with comas of various etiologies, as with hypoxia, drug related intoxication or deep anesthesia. The detection of bursts and the calculation of burst/suppression ratio are often used to monitor the level of anesthesia during treatment of status epilepticus. However, manual counting of bursts is a laborious process open to inter-rater variation and motivates a need for automatic detection. METHODS: We describe a novel unsupervised learning algorithm that detects bursts in EEG and generates burst-per-minute estimates for the purpose of monitoring sedation level in an intensive care unit (ICU). We validated the algorithm on 29 hours of burst annotated EEG data from 29 patients suffering from status epilepticus and hemorrhage. RESULTS: We report competitive results in comparison to neural networks learned via supervised learning. The mean absolute error (SD) in bursts per minute was 0.93 (1.38). CONCLUSION: We present a novel burst suppression detection algorithm that adapts to each patient individually, reports bursts-per-minute quickly, and does not require manual fine-tuning unlike previous approaches to burst-suppression pattern detection. SIGNIFICANCE: Our algorithm for automatic burst suppression quantification can greatly reduce manual oversight in depth of sedation monitoring.     

Synaptic damage underlies EEG abnormalities in postanoxic encephalopathy: A computational study

Objective

In postanoxic coma, EEG patterns indicate the severity of encephalopathy and typically evolve in time. We aim to improve the understanding of pathophysiological mechanisms underlying these EEG abnormalities.

Measuring the temporal dynamics of inter-personal neural entrainment in continuous child-adult EEG hyperscanning data

Current approaches to analysing EEG hyperscanning data in the developmental literature typically consider interpersonal entrainment between interacting physiological systems as a time-invariant property. This approach obscures crucial information about how entrainment between interacting systems is established and maintained over time. Here, we describe methods, and present computational algorithms, that will allow researchers to address this gap in the literature. We focus on how two different approaches to measuring entrainment, namely concurrent (e.g., power correlations, phase locking) and sequential (e.g., Granger causality) measures, can be applied to three aspects of the brain signal: amplitude, power, and phase. We guide the reader through worked examples using simulated data on how to leverage these methods to measure changes in interbrain entrainment. For each, we aim to provide a detailed explanation of the interpretation and application of these analyses when studying neural entrainment during early social interactions.     

Investigating the temporal dynamics of electroencephalogram (EEG) microstates using recurrent neural networks

Electroencephalogram (EEG) microstates that represent quasi‐stable, global neuronal activity are considered as the building blocks of brain dynamics. Therefore, the analysis of microstate sequences is a promising approach to understand fast brain dynamics that underlie various mental processes. Recent studies suggest that EEG microstate sequences are non‐Markovian and nonstationary, highlighting the importance of the sequential flow of information between different brain states. These findings inspired us to model these sequences using Recurrent Neural Networks (RNNs) consisting of long‐short‐term‐memory (LSTM) units to capture the complex temporal dependencies. Using an LSTM‐based auto encoder framework and different encoding schemes, we modeled the microstate sequences at multiple time scales (200–2,000 ms) aiming to capture stably recurring microstate patterns within and across subjects. We show that RNNs can learn underlying microstate patterns with high accuracy and that the microstate trajectories are subject invariant at shorter time scales (≤400 ms) and reproducible across sessions. Significant drop in the reconstruction accuracy was observed for longer sequence lengths of 2,000 ms. These findings indirectly corroborate earlier studies which indicated that EEG microstate sequences exhibit long‐range dependencies with finite memory content. Furthermore, we find that the latent representations learned by the RNNs are sensitive to external stimulation such as stress while the conventional univariate microstate measures (e.g., occurrence, mean duration, etc.) fail to capture such changes in brain dynamics. While RNNs cannot be configured to identify the specific discriminating patterns, they have the potential for learning the underlying temporal dynamics and are sensitive to sequence aberrations characterized by changes in metal processes. Empowered with the macroscopic understanding of the temporal dynamics that extends beyond short‐term interactions, RNNs offer a reliable alternative for exploring system level brain dynamics using EEG microstate sequences.     

Theta EEG dynamics of the error-related negativity.

OBJECTIVE: The error-related negativity (ERN) is a response-locked brain potential (ERP) occurring 80-100ms following response errors. This report contrasts three views of the genesis of the ERN, testing the classic view that time-locked phasic bursts give rise to the ERN against the view that the ERN arises from a pure phase-resetting of ongoing theta (4-7Hz) EEG activity and the view that the ERN is generated - at least in part - by a phase-resetting and amplitude enhancement of ongoing theta EEG activity. METHODS: Time-domain ERP analyses were augmented with time-frequency investigations of phase-locked and non-phase-locked spectral power, and inter-trial phase coherence (ITPC) computed from individual EEG trials, examining time courses and scalp topographies. Simulations based on the assumptions of the classic, pure phase-resetting, and phase-resetting plus enhancement views, using parameters from each subject's empirical data, were used to contrast the time-frequency findings that could be expected if one or more of these hypotheses adequately modeled the data. RESULTS: Error responses produced larger amplitude activity than correct responses in time-domain ERPs immediately following responses, as expected. Time-frequency analyses revealed that significant error-related post-response increases in total spectral power (phase- and non-phase-locked), phase-locked power, and ITPC were primarily restricted to the theta range, with this effect located over midfrontocentral sites, with a temporal distribution from approximately 150-200ms prior to the button press and persisting up to 400ms post-button press. The increase in non-phase-locked power (total power minus phase-locked power) was larger than phase-locked power, indicating that the bulk of the theta event-related dynamics were not phase-locked to response. Results of the simulations revealed a good fit for data simulated according to the phase-locking with amplitude enhancement perspective, and a poor fit for data simulated according to the classic view and the pure phase-resetting view. CONCLUSIONS: Error responses produce not only phase-locked increases in theta EEG activity, but also increases in non-phase-locked theta, both of which share a similar topography. SIGNIFICANCE: The findings are thus consistent with the notion advanced by Luu et al. [Luu P, Tucker DM, Makeig S. Frontal midline theta and the error-related negativity; neurophysiological mechanisms of action regulation. Clin Neurophysiol 2004;115:1821-35] that the ERN emerges, at least in part, from a phase-resetting and phase-locking of ongoing theta-band activity, in the context of a general increase in theta power following errors.     

Interictal EEG in temporal lobe epilepsy in childhood.

The authors clarified the value of interictal discharges and verified which extratemporal regions may also show epileptiform activity in temporal lobe epilepsy (TLE) in childhood. Thirty consecutive patients aged 3 to 18 years (mean age = 12.16 years; 16 male) with TLE associated with hippocampal atrophy were studied. Each patient had 1 to 15 interictal EEG recordings (mean: 5.6; total = 192 EEGs). Video-EEG monitoring was performed in 20 patients. All patients had MRI. The findings were compared with a control group of 53 consecutive TLE adult outpatients with hippocampal atrophy. Each adult patient underwent 3 to 21 routine EEGs (mean: 10.67; total = 566). Interictal EEGs of children with TLE showed extratemporal epileptiform discharges more frequently than EEGs of adults with TLE. Frontal, parietal, and occipital discharges were more frequently seen in children (P < 0.05). These results suggest a close interaction between temporal and other cerebral regions in children with epilepsy and provide further evidence of the existence of neural networks.     

Burst-suppression with identical bursts: A distinct EEG pattern with poor outcome in postanoxic coma

ObjectiveTo assess the incidence, quantified EEG characteristics, and prognostic significance of “burst-suppression with identical bursts” and to discuss potential pathophysiological mechanisms.MethodsBurst-suppression EEGs were identified from a cohort of 101 comatose patients after cardiac arrest, and from our complete database of 9600 EEGs, since 2005. Patterns with and without identical bursts were classified visually by two observers. Of patients after cardiac arrest, outcomes were assessed at three and six months. Identical and non-identical burst-suppression patterns were compared for quantified EEG characteristics and clinical outcome. Cross correlation of burstshape was applied to the first 500 ms of each burst.ResultsOf 9701 EEGs, 240 showed burst-suppression, 22 with identical bursts. Identical bursts were observed in twenty (20%) of 101 comatose patients after cardiac arrest between a median of 12 and 36 h after the arrest, but not in the six patients with other pathology than cerebral ischemia, or the 183 with anesthesia induced burst suppression. Inter-observer agreement was 0.8 and disagreement always resulted from sampling error. Burst-suppression with identical bursts was always bilateral synchronous, amplitudes were higher (128 vs. 25 μV, p = 0.0001) and correlation coefficients of burstshapes were higher (95% >0.75 vs. 0% >0.75, p < 0.0001) than in burst-suppression without identical bursts. All twenty patients with identical bursts after cardiac arrest had a poor outcome versus 10 (36%) without identical bursts.Conclusion“Burst-suppression with identical bursts” is a distinct pathological EEG pattern, which in this series only occurred after diffuse cerebral ischemia and was invariably associated with poor outcome.SignificanceIn comatose patients after cardiac arrest, “burst-suppression with identical bursts” predicts a poor outcome with a high specificity.     

The temporal dynamics of the stress response

This paper summarises the available evidence that failure of defense mechanisms in (semi)-natural social groups of animals may lead to serious forms of stress pathology. Hence the study of social stress may provide animal models with a high face validity. However, most of the animal models of human stress-disorders have concentrated on the consequences of chronic exposure to stressors. The present paper considers recent data, indicating that a single experience with a major stressor in the form of social defeat may have long-term consequences ranging from hours to days and weeks. It seems that the experience of a major stressor sensitizes the animal to subsequent stressors. The consequences of these long-term temporal dynamics of the stress response to the development of stress-related disorders and stress-vulnerability are discussed.