EEG synchronization measures are early outcome predictors in comatose patients after cardiac arrest

Deep learning for EEG-based prognostication after cardiac arrest: from current research to future clinical applications

Outcome prognostication in comatose patients after cardiac arrest (CA) remains to date a challenge. The major determinant of clinical outcome is the post-hypoxic/ischemic encephalopathy. Electroencephalography (EEG) is routinely used to assess neural functions in comatose patients. Currently, EEG-based outcome prognosis relies on visual evaluation by medical experts, which is time consuming, prone to subjectivity, and oblivious to complex patterns. The field of deep learning has given rise to powerful algorithms for detecting patterns in large amounts of data. Analyzing EEG signals of coma patients with deep neural networks with the goal of assisting in outcome prognosis is therefore a natural application of these algorithms. Here, we provide the first narrative literature review on the use of deep learning for prognostication after CA. Existing studies show overall high performance in predicting outcome, relying either on spontaneous or on auditory evoked EEG signals. Moreover, the literature is concerned with algorithmic interpretability, and has shown that largely, deep neural networks base their decisions on clinically or neurophysiologically meaningful features. We conclude this review by discussing considerations that the fields of artificial intelligence and neurology will need to jointly address in the future, in order for deep learning algorithms to break the publication barrier, and to be integrated in clinical practice.

Quantitative analysis of early-stage EEG reactivity predicts awakening and recovery of consciousness in patients with severe brain injury

BACKGROUND

Decisions of withdrawal of life-sustaining therapy for patients with severe brain injury are often based on prognostic evaluations such as analysis of electroencephalography (EEG) reactivity (EEG-R). However, EEG-R usually relies on visual assessment, which requires neurophysiological expertise and is prone to inter-rater variability. We hypothesised that quantitative analysis of EEG-R obtained 3 days after patient admission can identify new markers of subsequent awakening and consciousness recovery.

METHODS

In this prospective observational study of patients with severe brain injury requiring mechanical ventilation, quantitative EEG-R was assessed using standard 11-lead EEG with frequency-based (power spectral density) and functional connectivity-based (phase-lag index) analyses. Associations between awakening in the intensive care unit (ICU) and reactivity to auditory and nociceptive stimulations were assessed with logistic regression. Secondary outcomes included in-ICU mortality and 3-month Coma Recovery Scale-Revised (CRS-R) score.

RESULTS

Of 116 patients, 86 (74%) awoke in the ICU. Among quantitative EEG-R markers, variation in phase-lag index connectivity in the delta frequency band after noise stimulation was associated with awakening (adjusted odds ratio=0.89, 95% confidence interval: 0.81-0.97, P=0.02 corrected for multiple tests), independently of age, baseline severity, and sedation. This new marker was independently associated with improved 3-month CRS-R (adjusted β=-0.16, standard error 0.075, P=0.048), but not with mortality (adjusted odds ratio=1.08, 95% CI: 0.99-1.18, P=0.10).

CONCLUSIONS

An early-stage quantitative EEG-R marker was independently associated with awakening and 3-month level of consciousness in patients with severe brain injury. This promising marker based on functional connectivity will need external validation before potential integration into a multimodal prognostic model.

Auditory stimulation and deep learning predict awakening from coma after cardiac arrest

Assessing the integrity of neural functions in coma after cardiac arrest remains an open challenge. Prognostication of coma outcome relies mainly on visual expert scoring of physiological signals, which is prone to subjectivity and leaves a considerable number of patients in a 'grey zone', with uncertain prognosis. Quantitative analysis of EEG responses to auditory stimuli can provide a window into neural functions in coma and information about patients' chances of awakening. However, responses to standardized auditory stimulation are far from being used in a clinical routine due to heterogeneous and cumbersome protocols. Here, we hypothesize that convolutional neural networks can assist in extracting interpretable patterns of EEG responses to auditory stimuli during the first day of coma that are predictive of patients' chances of awakening and survival at 3 months. We used convolutional neural networks (CNNs) to model single-trial EEG responses to auditory stimuli in the first day of coma, under standardized sedation and targeted temperature management, in a multicentre and multiprotocol patient cohort and predict outcome at 3 months. The use of CNNs resulted in a positive predictive power for predicting awakening of 0.83 ± 0.04 and 0.81 ± 0.06 and an area under the curve in predicting outcome of 0.69 ± 0.05 and 0.70 ± 0.05, for patients undergoing therapeutic hypothermia and normothermia, respectively. These results also persisted in a subset of patients that were in a clinical 'grey zone'. The network's confidence in predicting outcome was based on interpretable features: it strongly correlated to the neural synchrony and complexity of EEG responses and was modulated by independent clinical evaluations, such as the EEG reactivity, background burst-suppression or motor responses. Our results highlight the strong potential of interpretable deep learning algorithms in combination with auditory stimulation to improve prognostication of coma outcome.

Prognostication after cardiac arrest: how EEG and evoked potentials may improve the challenge

About 80% of patients resuscitated from CA are comatose at ICU admission and nearly 50% of survivors are still unawake at 72 h. Predicting neurological outcome of these patients is important to provide correct information to patient's relatives, avoid disproportionate care in patients with irreversible hypoxic-ischemic brain injury (HIBI) and inappropriate withdrawal of care in patients with a possible favorable neurological recovery. ERC/ESICM 2021 algorithm allows a classification as "poor outcome likely" in 32%, the outcome remaining "indeterminate" in 68%. The crucial question is to know how we could improve the assessment of both unfavorable but also favorable outcome prediction. Neurophysiological tests, i.e., electroencephalography (EEG) and evoked-potentials (EPs) are a non-invasive bedside investigations. The EEG is the record of brain electrical fields, characterized by a high temporal resolution but a low spatial resolution. EEG is largely available, and represented the most widely tool use in recent survey examining current neuro-prognostication practices. The severity of HIBI is correlated with the predominant frequency and background continuity of EEG leading to "highly malignant" patterns as suppression or burst suppression in the most severe HIBI. EPs differ from EEG signals as they are stimulus induced and represent the summated activities of large populations of neurons firing in synchrony, requiring the average of numerous stimulations. Different EPs (i.e., somato sensory EPs (SSEPs), brainstem auditory EPs (BAEPs), middle latency auditory EPs (MLAEPs) and long latency event-related potentials (ERPs) with mismatch negativity (MMN) and P300 responses) can be assessed in ICU, with different brain generators and prognostic values. In the present review, we summarize EEG and EPs signal generators, recording modalities, interpretation and prognostic values of these different neurophysiological tools. Finally, we assess the perspective for futures neurophysiological investigations, aiming to reduce prognostic uncertainty in comatose and disorders of consciousness (DoC) patients after CA.

Repetitive Electroencephalography as Biomarker for the Prediction of Survival in Patients with Post-Hypoxic Encephalopathy

Predicting survival in patients with post-hypoxic encephalopathy (HE) after cardiopulmonary resuscitation is a challenging aspect of modern neurocritical care. Here, continuous electroencephalography (cEEG) has been established as the gold standard for neurophysiological outcome prediction. Unfortunately, cEEG is not comprehensively available, especially in rural regions and developing countries. The objective of this monocentric study was to investigate the predictive properties of repetitive EEGs (rEEGs) with respect to 12-month survival based on data for 199 adult patients with HE, using log-rank and multivariate Cox regression analysis (MCRA). A total number of 59 patients (29.6%) received more than one EEG during the first 14 days of acute neurocritical care. These patients were analyzed for the presence of and changes in specific EEG patterns that have been shown to be associated with favorable or poor outcomes in HE. Based on MCRA, an initially normal amplitude with secondary low-voltage EEG remained as the only significant predictor for an unfavorable outcome, whereas all other relevant parameters identified by univariate analysis remained non-significant in the model. In conclusion, rEEG during early neurocritical care may help to assess the prognosis of HE patients if cEEG is not available.

Diagnostic and prognostic EEG analysis of critically ill patients: A deep learning study

Visual interpretation of electroencephalography (EEG) is time consuming, may lack objectivity, and is restricted to features detectable by a human. Computer-based approaches, especially deep learning, could potentially overcome these limitations. However, most deep learning studies focus on a specific question or a single pathology. Here we explore the potential of deep learning for EEG-based diagnostic and prognostic assessment of patients with acute consciousness impairment (ACI) of various etiologies. EEGs from 358 adults from a randomized controlled trial (CERTA, NCT03129438) were retrospectively analyzed. A convolutional neural network was used to predict the clinical outcome (based either on survival or on best cerebral performance category) and to determine the etiology (four diagnostic categories). The largest probability output served as marker for the confidence of the network in its prediction ("certainty factor"); we also systematically compared the predictions with raw EEG data, and used a visualization algorithm (Grad-CAM) to highlight discriminative patterns. When all patients were considered, the area under the receiver operating characteristic curve (AUC) was 0.721 for predicting survival and 0.703 for predicting the outcome based on best CPC; for patients with certainty factor ≥ 60 % the AUCs increased to 0.776 and 0.755 respectively; and for certainty factor ≥ 75 % to 0.852 and 0.879. The accuracy for predicting the etiology was 54.5 %; the accuracy increased to 67.7 %, 70.3 % and 84.1 % for patients with certainty factor of 50 %, 60 % and 75 % respectively. Visual analysis showed that the network learnt EEG patterns typically recognized by human experts, and suggested new criteria. This work demonstrates for the first time the potential of deep learning-based EEG analysis in critically ill patients with various etiologies of ACI. Certainty factor and post-hoc correlation of input data with prediction help to better characterize the method and pave the route for future implementations in clinical routine.

Comatose Patients After Cardiopulmonary Resuscitation: An Analysis Based on Quantitative Methods of EEG Reactivity

Objective

Every year, approximately 50–110/1,00,000 people worldwide suffer from cardiac arrest, followed by hypoxic-ischemic encephalopathy after cardiopulmonary resuscitation (CPR), and approximately 40–66% of patients do not recover. The purpose of this study was to identify the brain network parameters and key brain regions associated with awakening by comparing the reactivity characteristics of the brain networks between the awakening and unawakening groups of CPR patients after coma, thereby providing a basis for further awakening interventions.

Method

This study involved a prospective cohort study. Using a 64-electrode electroencephalography (EEG) wireless 64A system, EEG signals were recorded from 16 comatose patients after CPR in the acute phase (<1 month) from 2019 to 2020. MATLAB (2017b) was used to quantitatively analyze the reactivity (power spectrum and entropy) and brain network characteristics (coherence and phase lag index) after pain stimulation. The patients were divided into an awakening group and an unawakening group based on their ability to execute commands or engage in repeated and continuous purposeful behavior after 3 months. The above parameters were compared to determine whether there were differences between the two groups.

Results

(1) Power spectrum: the awakening group had higher gamma, beta and alpha spectral power after pain stimulation in the frontal and parietal lobes, and lower delta and theta spectral power in the bilateral temporal and occipital lobes than the unawakening group. (2) Entropy: after pain stimulation, the awakening group had higher entropy in the frontal and parietal lobes and lower entropy in the temporal occipital lobes than the unawakening group. (3) Connectivity: after pain stimulation, the awakening group had stronger gamma and beta connectivity in nearly the whole brain, but weaker theta and delta connectivity in some brain regions (e.g., the frontal-occipital lobe and parietal-occipital lobe) than the unawakening group.

Conclusion

After CPR, comatose patients were more likely to awaken if there was a higher stimulation of fast-frequency band spectral power, higher entropy, stronger whole-brain connectivity and better retention of frontal-parietal lobe function after pain stimulation.

Complementary roles of neural synchrony and complexity for indexing consciousness and chances of surviving in acute coma

An open challenge in consciousness research is understanding how neural functions are altered by pathological loss of consciousness. To maintain consciousness, the brain needs synchronized communication of information across brain regions, and sufficient complexity in neural activity. Coordination of brain activity, typically indexed through measures of neural synchrony, has been shown to decrease when consciousness is lost and to reflect the clinical state of patients with disorders of consciousness. Moreover, when consciousness is lost, neural activity loses complexity, while the levels of neural noise, indexed by the slope of the electroencephalography (EEG) spectral exponent decrease. Although these properties have been well investigated in resting state activity, it remains unknown whether the sensory processing network, which has been shown to be preserved in coma, suffers from a loss of synchronization or information content. Here, we focused on acute coma and hypothesized that neural synchrony in response to auditory stimuli would reflect coma severity, while complexity, or neural noise, would reflect the presence or loss of consciousness. Results showed that neural synchrony of EEG signals was stronger for survivors than non-survivors and predictive of patients' outcome, but indistinguishable between survivors and healthy controls. Measures of neural complexity and neural noise were not informative of patients' outcome and had high or low values for patients compared to controls. Our results suggest different roles for neural synchrony and complexity in acute coma. Synchrony represents a precondition for consciousness, while complexity needs an equilibrium between high or low values to support conscious cognition.

EEG patterns and their correlations with short- and long-term mortality in patients with hypoxic encephalopathy

OBJECTIVE

To analyze the association between electroencephalographic (EEG) patterns and overall, short- and long-term mortality in patients with hypoxic encephalopathy (HE).

METHODS

Retrospective, mono-center analysis of 199 patients using univariate log-rank tests (LR) and multivariate cox regression (MCR).

RESULTS

Short-term mortality, defined as death within 30-days post-discharge was 54.8%. Long-term mortality rates were 69.8%, 71.9%, and 72.9%, at 12-, 24-, and 36-months post-HE, respectively. LR revealed a significant association between EEG suppression (SUP) and short-term mortality, and identified low voltage EEG (LV), burst suppression (BSP), periodic discharges (PD) and post-hypoxic status epilepticus (PSE) as well as missing (aBA) or non-reactive background activity (nrBA) as predictors for overall, short- and long-term mortality. MCR indicated SUP, LV, BSP, PD, aBA and nrBA as significantly associated with overall and short-term mortality to varying extents. LV and BSP were significant predictors for long-term mortality in short-term survivors. Rhythmic delta activity, stimulus induced rhythmic, periodic or ictal discharges and sharp waves were not significantly associated with a higher mortality.

CONCLUSION

The presence of several specific EEG patterns can help to predict overall, short- and long-term mortality in HE patients.

SIGNIFICANCE

The present findings may help to improve the challenging prognosis estimation in HE patients.

EEG functional connectivity contributes to outcome prediction of postanoxic coma

OBJECTIVE

To 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.

METHODS

Prospective, 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).

RESULTS

We 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.

CONCLUSION

Functional connectivity measures improve EEG based prediction models for poor outcome of postanoxic coma.

SIGNIFICANCE

Functional connectivity features derived from early EEG hold potential to improve outcome prediction of coma after cardiac arrest.

Standardized visual EEG features predict outcome in patients with acute consciousness impairment of various etiologies

BACKGROUND

Early prognostication in patients with acute consciousness impairment is a challenging but essential task. Current prognostic guidelines vary with the underlying etiology. In particular, electroencephalography (EEG) is the most important paraclinical examination tool in patients with hypoxic ischemic encephalopathy (HIE), whereas it is not routinely used for outcome prediction in patients with traumatic brain injury (TBI).

METHOD

Data from 364 critically ill patients with acute consciousness impairment (GCS ≤ 11 or FOUR ≤ 12) of various etiologies and without recent signs of seizures from a prospective randomized trial were retrospectively analyzed. Random forest classifiers were trained using 8 visual EEG features-first alone, then in combination with clinical features-to predict survival at 6 months or favorable functional outcome (defined as cerebral performance category 1-2).

RESULTS

The area under the ROC curve was 0.812 for predicting survival and 0.790 for predicting favorable outcome using EEG features. Adding clinical features did not improve the overall performance of the classifier (for survival: AUC = 0.806, p = 0.926; for favorable outcome: AUC = 0.777, p = 0.844). Survival could be predicted in all etiology groups: the AUC was 0.958 for patients with HIE, 0.955 for patients with TBI and other neurosurgical diagnoses, 0.697 for patients with metabolic, inflammatory or infectious causes for consciousness impairment and 0.695 for patients with stroke. Training the classifier separately on subgroups of patients with a given etiology (and thus using less training data) leads to poorer classification performance.

CONCLUSIONS

While prognostication was best for patients with HIE and TBI, our study demonstrates that similar EEG criteria can be used in patients with various causes of consciousness impairment, and that the size of the training set is more important than homogeneity of ACI etiology.

Quantitative Electroencephalogram Trends Predict Recovery in Hypoxic-Ischemic Encephalopathy

OBJECTIVES

Electroencephalogram features predict neurologic recovery following cardiac arrest. Recent work has shown that prognostic implications of some key electroencephalogram features change over time. We explore whether time dependence exists for an expanded selection of quantitative electroencephalogram features and whether accounting for this time dependence enables better prognostic predictions.

DESIGN

Retrospective.

SETTING

ICUs at four academic medical centers in the United States.

PATIENTS

Comatose patients with acute hypoxic-ischemic encephalopathy.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

We analyzed 12,397 hours of electroencephalogram from 438 subjects. From the electroencephalogram, we extracted 52 features that quantify signal complexity, category, and connectivity. We modeled associations between dichotomized neurologic outcome (good vs poor) and quantitative electroencephalogram features in 12-hour intervals using sequential logistic regression with Elastic Net regularization. We compared a predictive model using time-varying features to a model using time-invariant features and to models based on two prior published approaches. Models were evaluated for their ability to predict binary outcomes using area under the receiver operator curve, model calibration (how closely the predicted probability of good outcomes matches the observed proportion of good outcomes), and sensitivity at several common specificity thresholds of interest. A model using time-dependent features outperformed (area under the receiver operator curve, 0.83 ± 0.08) one trained with time-invariant features (0.79 ± 0.07; p < 0.05) and a random forest approach (0.74 ± 0.13; p < 0.05). The time-sensitive model was also the best-calibrated.

CONCLUSIONS

The statistical association between quantitative electroencephalogram features and neurologic outcome changed over time, and accounting for these changes improved prognostication performance.

Electrographic predictors of successful weaning from anaesthetics in refractory status epilepticus

Intravenous third-line anaesthetic agents are typically titrated in refractory status epilepticus to achieve either seizure suppression or burst suppression on continuous EEG. However, the optimum treatment paradigm is unknown and little data exist to guide the withdrawal of anaesthetics in refractory status epilepticus. Premature withdrawal of anaesthetics risks the recurrence of seizures, whereas the prolonged use of anaesthetics increases the risk of treatment-associated adverse effects. This study sought to measure the accuracy of features of EEG activity during anaesthetic weaning in refractory status epilepticus as predictors of successful weaning from intravenous anaesthetics. We prespecified a successful anaesthetic wean as the discontinuation of intravenous anaesthesia without developing recurrent status epilepticus, and a wean failure as either recurrent status epilepticus or the resumption of anaesthesia for the purpose of treating an EEG pattern concerning for incipient status epilepticus. We evaluated two types of features as predictors of successful weaning: spectral components of the EEG signal, and spatial-correlation-based measures of functional connectivity. The results of these analyses were used to train a classifier to predict wean outcome. Forty-seven consecutive anaesthetic weans (23 successes, 24 failures) were identified from a single-centre cohort of patients admitted with refractory status epilepticus from 2016 to 2019. Spectral components of the EEG revealed no significant differences between successful and unsuccessful weans. Analysis of functional connectivity measures revealed that successful anaesthetic weans were characterized by the emergence of larger, more densely connected, and more highly clustered spatial functional networks, yielding 75.5% (95% confidence interval: 73.1-77.8%) testing accuracy in a bootstrap analysis using a hold-out sample of 20% of data for testing and 74.6% (95% confidence interval 73.2-75.9%) testing accuracy in a secondary external validation cohort, with an area under the curve of 83.3%. Distinct signatures in the spatial networks of functional connectivity emerge during successful anaesthetic liberation in status epilepticus; these findings are absent in patients with anaesthetic wean failure. Identifying features that emerge during successful anaesthetic weaning may allow faster and more successful anaesthetic liberation after refractory status epilepticus.

EEG-based outcome prediction after cardiac arrest with convolutional neural networks: Performance and visualization of discriminative features

Prognostication for comatose patients after cardiac arrest is a difficult but essential task. Currently, visual interpretation of electroencephalogram (EEG) is one of the main modality used in outcome prediction. There is a growing interest in computer-assisted EEG interpretation, either to overcome the possible subjectivity of visual interpretation, or to identify complex features of the EEG signal. We used a one-dimensional convolutional neural network (CNN) to predict functional outcome based on 19-channel-EEG recorded from 267 adult comatose patients during targeted temperature management after CA. The area under the receiver operating characteristic curve (AUC) on the test set was 0.885. Interestingly, model architecture and fine-tuning only played a marginal role in classification performance. We then used gradient-weighted class activation mapping (Grad-CAM) as visualization technique to identify which EEG features were used by the network to classify an EEG epoch as favorable or unfavorable outcome, and also to understand failures of the network. Grad-CAM showed that the network relied on similar features than classical visual analysis for predicting unfavorable outcome (suppressed background, epileptiform transients). This study confirms that CNNs are promising models for EEG-based prognostication in comatose patients, and that Grad-CAM can provide explanation for the models' decision-making, which is of utmost importance for future use of deep learning models in a clinical setting.

Electroencephalography-based power spectra allow coma outcome prediction within 24 h of cardiac arrest

BACKGROUND

Outcome prediction in comatose patients following cardiac arrest remains challenging. Here, we assess the predictive performance of electroencephalography-based power spectra within 24 h from coma onset.

METHODS

We acquired electroencephalography (EEG) from comatose patients (n = 138) on the first day of coma in four hospital sites in Switzerland. Outcome was categorised as favourable or unfavourable based on the best state within three months. Data were split in training and test sets. We evaluated the predictive performance of EEG power spectra for long term outcome and its added value to standard clinical tests.

RESULTS

Out of 138 patients, 80 had a favourable outcome. Power spectra comparison between favourable and unfavourable outcome in the training set yielded significant differences at 5.2-13.2 Hz and above 21 Hz. Outcome prediction based on power at 5.2-13.2 Hz was accurate in training and test sets. Overall, power spectra predicted patients' outcome with maximum specificity and positive predictive value: 1.00 (95% with CI: 0.94-1.00 and 0.89-1.00, respectively). The combination of power spectra and reactivity yielded better accuracy and sensitivity (0.81, 95% CI: 0.71-0.89) than prediction based on power spectra alone.

CONCLUSIONS

On the first day of coma following cardiac arrest, low power spectra values around 10 Hz, typically linked to impaired cortico-thalamic structural connections, are highly specific of unfavourable outcome. Peaks in this frequency range can predict long-term outcome.

Global Epileptic Seizure Identification With Affinity Propagation Clustering Partition Mutual Information Using Cross-Layer Fully Connected Neural Network

A longstanding challenge in epilepsy research and practice is the need to classify synchronization patterns hidden in multivariate electroencephalography (EEG) data that is routinely superimposed with intensive noise. It is essential to select a suitable feature extraction method to achieve high recognition performance. A typical approach is to extract the mutual information (MI) between pairs of channels. This calculation, which considers the differences between the sequence pairs to build a reasonable partition, can improve the classification performance. On this basis, however, it is even more difficult to adaptively classify the synchronization patterns hidden in multivariate EEG data under circumstances of insufficient a priori knowledge of domain dependency, such as denoising, feature extraction on a special patient, etc. To address these problems by (1) effectively calculating the MI matrix (synchronization pattern) and (2) accurately classifying the seizure or non-seizure state, this study first accurately measures the synchronization between channel pairs in terms of affinity propagation clustering partition MI (APCPMI). The global synchronization measurement is then obtained by organizing APCPMIs of all channel pairs into a correlation matrix. Finally, a cross-layer fully connected net is designed to characterize the synchronization dynamics correlation matrices adaptively and identify seizure or non-seizure states automatically. Experiments are performed using the CHB-MIT scalp EEG dataset to evaluate the proposed approach. Seizure states are identified with an accuracy, sensitivity, and specificity of 0.9793 ± 0.002, 0.9942 ± 0.0005, and 0.9676 ± 0.003, respectively; the resulting performance is superior to those achieved by most existing methods over the same dataset. Furthermore, the approach alleviates the necessity for strictly preprocessing (denoising, removing interferences and artifacts) the EEG data using prior knowledge, which is usually required by existing approaches.

Predictive value of EEG in postanoxic encephalopathy: A quantitative model-based approach

INTRODUCTION

The majority of comatose patients after cardiac arrest do not regain consciousness due to severe postanoxic encephalopathy. Early and accurate outcome prediction is therefore essential in determining further therapeutic interventions. The electroencephalogram is a standardized and commonly available tool used to estimate prognosis in postanoxic patients. The identification of pathological EEG patterns with poor prognosis relies however primarily on visual EEG scoring by experts. We introduced a model-based approach of EEG analysis (state space model) that allows for an objective and quantitative description of spectral EEG variability.

METHODS

We retrospectively analyzed standard EEG recordings in 83 comatose patients after cardiac arrest between 2005 and 2013 in the intensive care unit of the University Hospital Zürich. Neurological outcome was assessed one month after cardiac arrest using the Cerebral Performance Category. For a dynamic and quantitative EEG analysis, we implemented a model-based approach (state space analysis) to quantify EEG background variability independent from visual scoring of EEG epochs. Spectral variability was compared between groups and correlated with clinical outcome parameters and visual EEG patterns.

RESULTS

Quantitative assessment of spectral EEG variability (state space velocity) revealed significant differences between patients with poor and good outcome after cardiac arrest: Lower mean velocity in temporal electrodes (T4 and T5) was significantly associated with poor prognostic outcome (p<0.005) and correlated with independently identified visual EEG patterns such as generalized periodic discharges (p<0.02). Receiver operating characteristic (ROC) analysis confirmed the predictive value of lower state space velocity for poor clinical outcome after cardiac arrest (AUC 80.8, 70% sensitivity, 15% false positive rate).

CONCLUSION

Model-based quantitative EEG analysis (state space analysis) provides a novel, complementary marker for prognosis in postanoxic encephalopathy.