Continuous EEG Monitoring in the Intensive Care Unit

Calibration trial of an innovative medical device (NEVVA© ) for the evaluation of pain in non-communicating patients in the intensive care unit

Background

Pain management is an essential and complex issue for non-communicative patients undergoing sedation in the intensive care unit (ICU). The Behavioral Pain Scale (BPS), although not perfect for assessing behavioral pain, is the gold standard based partly on clinical facial expression. NEVVA© , an automatic pain assessment tool based on facial expressions in critically ill patients, is a much-needed innovative medical device.

Methods

In this prospective pilot study, we recorded the facial expressions of critically ill patients in the medical ICU of Caen University Hospital using the iPhone and Smart Motion Tracking System (SMTS) software with the Facial Action Coding System (FACS) to measure human facial expressions metrically during sedation weaning. Analyses were recorded continuously, and BPS scores were collected hourly over two 8 h periods per day for 3 consecutive days. For this first stage, calibration of the innovative NEVVA© medical device algorithm was obtained by comparison with the reference pain scale (BPS).

Results

Thirty participants were enrolled between March and July 2022. To assess the acute severity of illness, the Sequential Organ Failure Assessment (SOFA) and the Simplified Acute Physiology Score (SAPS II) were recorded on ICU admission and were 9 and 47, respectively. All participants had deep sedation, assessed by a Richmond Agitation and Sedation scale (RASS) score of less than or equal to -4 at the time of inclusion. One thousand and six BPS recordings were obtained, and 130 recordings were retained for final calibration: 108 BPS recordings corresponding to the absence of pain and 22 BPS recordings corresponding to the presence of pain. Due to the small size of the dataset, a leave-one-subject-out cross-validation (LOSO-CV) strategy was performed, and the training results obtained the receiver operating characteristic (ROC) curve with an area under the curve (AUC) of 0.792. This model has a sensitivity of 81.8% and a specificity of 72.2%.

Conclusion

This pilot study calibrated the NEVVA© medical device and showed the feasibility of continuous facial expression analysis for pain monitoring in ICU patients. The next step will be to correlate this device with the BPS scale.

A TDM-Based 16-Channel AFE ASIC With Enhanced System-Level CMRR for Wearable EEG Recording With Dry Electrodes

A multi-channel analog front-end (AFE) ASIC for wearable EEG recording application is presented in this article. Two techniques, namely chopping stabilization (CS) and time-division-multiplexing (TDM) are combined in a unified manner to improve the input-referred noise and the system level common-mode rejection ratio (CMRR) for multi-channel AFE. With the proposed TDM/CS structure, multiple channels can share single second-stage amplifier for significant reduction in chip size and power consumption. Dual feedback loops for input impedance boosting as well as electrode offset cancellation are incorporated in the system. Implemented in a 0.18-μm CMOS process, the AFE consumes 24 μW under 1 V supply. The input referred noise is 0.63 μV_rms in 0.5 Hz-100 Hz and the input impedance is boosted to 560 MΩ at 50 Hz. The measured amplifier intrinsic CMRR and system-level AFE CMRR are 89 dB and 82 dB, respectively.

Rapid Annotation of Seizures and Interictal-ictal Continuum EEG Patterns

Seizures, status epilepticus, and seizure-like rhythmic or periodic activities are common, pathological, harmful states of brain electrical activity seen in the electroencephalogram (EEG) of patients during critical medical illnesses or acute brain injury. Accumulating evidence shows that these states, when prolonged, cause neurological injury. In this study we developed a valid method to automatically discover a small number of homogeneous pattern clusters, to facilitate efficient interactive labelling by EEG experts. 592 time domain and spectral features were extracted from continuous EEG (cEEG) data of 369 ICU (intensive care unit) patients. For each patient, feature dimensionality was reduced using principal component analysis (PCA), retaining 95% of the variance. K-medoids clustering was applied to learn a local dictionary from each patient, consisting of k=100 exemplars/words. Changepoint detection (CPD) was utilized to break each EEG into segments. A bag-of-words (BoW) representation was computed for each segment, specifically, a normalized histogram of the words found within each segment. Segments were further clustered using the BoW histograms by Affinity Propagation (AP) using a χ² distance to measure similarities between histograms. The resulting 30 50 clusters for each patient were scored by EEG experts through labeling only the cluster medoids. Embedding methods t-SNE (t-distributed stochastic neighbor embedding) and PCA were used to provide a 2D representation for visualization and exploration of the data. Our results illustrate that it takes approximately 3 minutes to annotate 24 hours of cEEG by experts, which is at least 60 times faster than unaided manual review.

Prevention, Treatment, and Monitoring of Seizures in the Intensive Care Unit

The diagnosis and management of seizures in the critically ill patient can sometimes present a unique challenge for practitioners due to lack of exposure and complex patient comorbidities. The reported incidence varies between 8% and 34% of critically ill patients, with many patients often showing no overt clinical signs of seizures. Outcomes in patients with unidentified seizure activity tend to be poor, and mortality significantly increases in those who have seizure activity longer than 30 min. Prompt diagnosis and provision of medical therapy are crucial in order to attain successful seizure termination and prevent poor outcomes. In this article, we review the epidemiology and pathophysiology of seizures in the critically ill, various seizure monitoring modalities, and recommended medical therapy.

Prescription patterns for routine EEG ordering in patients with intracranial hemorrhage admitted to a neurointensive care unit

PURPOSE

To examine clinical factors, including established electroencephalography (EEG) consensus recommendations, that may influence EEG-prescription in critically-ill intracerebral hemorrhage (ICH) patients in the neurointensive care unit.

METHODS

Retrospective analysis of 330 ICH patients admitted to a neurointensive care unit at an academic medical center between 01/2013-12/2015. We compared EEG prescription patterns with current EEG consensus recommendations, and employed univariate and multivariable logistic regression modeling to determine clinical variables associated with EEG ordering.

RESULTS

Seventy-eight (41%) of 190 subjects underwent EEG in accordance with EEG-consensus guidelines, demonstrating an overall accuracy (probability that EEG prescription aligned with EEG consensus recommendations) of 64.6% (95%-CI59.1-69.7). Factors independently associated with EEG ordering included fulfillment of EEG consensus recommendations, lower admission Glasgow Coma Scale (GCS), and presence of clinical seizures. The unadjusted and adjusted C-statistics for fulfillment of consensus recommendations was 0.74 (95%-CI 0.69-0.80) and 0.85 (95%-CI 0.81-0.90), respectively. Among 83 subjects undergoing EEG (25.2%), EEG findings informed clinical decision-making in 50 patients (60%).

CONCLUSIONS

EEG appeared underused in ICH, since <50% of patients who fulfilled guideline criteria underwent EEG. Prescription of EEG was related to factors beyond those included in consensus recommendations. Validation of our findings and their association with outcome is required.

Value and mechanisms of EEG reactivity in the prognosis of patients with impaired consciousness: a systematic review

Background

Electroencephalography (EEG) is a well-established tool for assessing brain function that is available at the bedside in the intensive care unit (ICU). This review aims to discuss the relevance of electroencephalographic reactivity (EEG-R) in patients with impaired consciousness and to describe the neurophysiological mechanisms involved.

Methods

We conducted a systematic search of the term “EEG reactivity and coma” using the PubMed database. The search encompassed articles published from inception to March 2018 and produced 202 articles, of which 42 were deemed relevant, assessing the importance of EEG-R in relationship to outcomes in patients with impaired consciousness, and were therefore included in this review.

Results

Although definitions, characteristics and methods used to assess EEG-R are heterogeneous, several studies underline that a lack of EEG-R is associated with mortality and unfavorable outcome in patients with impaired consciousness. However, preserved EEG-R is linked to better odds of survival. Exploring EEG-R to nociceptive, auditory, and visual stimuli enables a noninvasive trimodal functional assessment of peripheral and central sensory ascending pathways that project to the brainstem, the thalamus and the cerebral cortex. A lack of EEG-R in patients with impaired consciousness may result from altered modulation of thalamocortical loop activity by afferent sensory input due to neural impairment. Assessing EEG-R is a valuable tool for the diagnosis and outcome prediction of severe brain dysfunction in critically ill patients.

Conclusions

This review emphasizes that whatever the etiology, patients with impaired consciousness featuring a reactive electroencephalogram are more likely to have a favorable outcome, whereas those with a nonreactive electroencephalogram are prone to having an unfavorable outcome. EEG-R is therefore a valuable prognostic parameter and warrants a rigorous assessment. However, current assessment methods are heterogeneous, and no consensus exists. Standardization of stimulation and interpretation methods is needed.

Neurophysiological assessment of brain dysfunction in critically ill patients: an update

The aim of this review was to provide up-to-date information about the usefulness of clinical neurophysiology testing in the management of critically ill patients. Evoked potentials (EPs) and electroencephalogram (EEG) are non-invasive clinical neurophysiology tools that allow an objective assessment of the central nervous system's function at the bedside in intensive care unit (ICU). These tests are quite useful in diagnosing cerebral complications, and establishing the vital and functional prognosis in ICU. EEG keeps a particularly privileged importance in detecting seizures phenomena such as subclinical seizures and non-convulsive status epilepticus. Quantitative EEG (QEEG) analysis techniques commonly called EEG Brain mapping can provide obvious topographic displays of digital EEG signals characteristics, showing the potential distribution over the entire scalp including filtering, frequency, and amplitude analysis and color mapping. Evidences of usefulness of QEEG for seizures detection in ICU are provided by several recent studies. Furthermore, beyond detection of epileptic phenomena, changes of some QEEG panels are early warning indicators of sedation level as well as brain damage or dysfunction in ICU. EPs offer the opportunity for assessing brainstem's functional integrity, as well as subcortical and cortical brain areas. A multimodal use, combining EEG and various modalities of EPs is recommended since this allows a more accurate functional exploration of the brain and helps caregivers to tailor therapeutic measures according to neurological worsening trends and to anticipate the prognosis in ICU.

Integration of EEG lead placement templates into traditional technologist-based staffing models reduces costs in continuous video-EEG monitoring service

PURPOSE

The purpose of this study was to determine the relative cost reductions within different staffing models for continuous video-electroencephalography (cvEEG) service by introducing a template system for 10/20 lead application.

METHODS

We compared six staffing models using decision tree modeling based on historical service line utilization data from the cvEEG service at our center. Templates were integrated into technologist-based service lines in six different ways. The six models studied were templates for all studies, templates for intensive care unit (ICU) studies, templates for on-call studies, templates for studies of ≤ 24-hour duration, technologists for on-call studies, and technologists for all studies.

RESULTS

Cost was linearly related to the study volume for all models with the "templates for all" model incurring the lowest cost. The "technologists for all" model carried the greatest cost. Direct cost comparison shows that any introduction of templates results in cost savings, with the templates being used for patients located in the ICU being the second most cost efficient and the most practical of the combined models to implement. Cost difference between the highest and lowest cost models under the base case produced an annual estimated savings of $267,574. Implementation of the ICU template model at our institution under base case conditions would result in a $205,230 savings over our current "technologist for all" model.

CONCLUSIONS

Any implementation of templates into a technologist-based cvEEG service line results in cost savings, with the most significant annual savings coming from using the templates for all studies, but the most practical implementation approach with the second highest cost reduction being the template used in the ICU. The lowered costs determined in this work suggest that a template-based cvEEG service could be supported at smaller centers with significantly reduced costs and could allow for broader use of cvEEG patient monitoring.

Inter-rater agreement on identification of electrographic seizures and periodic discharges in ICU EEG recordings

OBJECTIVE

This study investigated inter-rater agreement (IRA) among EEG experts for the identification of electrographic seizures and periodic discharges (PDs) in continuous ICU EEG recordings.

METHODS

Eight board-certified EEG experts independently identified seizures and PDs in thirty 1-h EEG segments which were selected from ICU EEG recordings collected from three medical centers. IRA was compared between seizure and PD identifications, as well as among rater groups that have passed an ICU EEG Certification Test, developed by the Critical Care EEG Monitoring Research Consortium (CCEMRC).

RESULTS

Both kappa and event-based IRA statistics showed higher mean values in identification of seizures compared to PDs (k=0.58 vs. 0.38; p<0.001). The group of rater pairs who had both passed the ICU EEG Certification Test had a significantly higher mean IRA in comparison to rater pairs in which neither had passed the test.

CONCLUSIONS

IRA among experts is significantly higher for identification of electrographic seizures compared to PDs. Additional instruction, such as the training module and certification test developed by the CCEMRC, could enhance this IRA.

SIGNIFICANCE

This study demonstrates more disagreement in the labeling of PDs in comparison to seizures. This may be improved by education about standard EEG nomenclature.

Electroencephalography leads placed by nontechnologists using a template system produce signals equal in quality to technologist-applied, collodion disk leads

The purpose of this study was to compare the quality of the electroencephalographic (EEG) data obtained with a BraiNet template in a practical use setting, to that obtained with standard 10/20 spaced, technologist-applied, collodion-based disk leads. Pairs of 8-hour blocks of EEG data were prospectively collected from 32 patients with a Glasgow coma score of ≤9 and clinical concern for underlying nonconvulsive status epilepticus over a 6-month period in the Neurocritical Care Unit at the Duke University Medical Center. The studies were initiated with the BraiNet template system applied by critical care nurse practitioners or physicians, followed by standard, collodion leads applied by registered technologists using the 10/20 system of placement. Impedances were measured at the beginning and end of each block recorded and variance in impedance, mean impedance, and the largest differences in impedances found within a given lead set were compared. Physicians experienced in reading EEG performed a masked review of the EEG segments obtained to assess the subjective quality of the recordings obtained with the templates. We found no clinically significant differences in the impedance measures. There was a 3-hour reduction in the time required to initiate EEG recording using the templates (P < 0.001). There was no difference in the overall subjective quality distributions for template-applied versus technologist-applied EEG leads. The templates were also found to be well accepted by the primary users in the intensive care unit. The findings suggest that the EEG data obtained with this approach are comparable with that obtained by registered technologist-applied leads and represents a possible solution to the growing clinical need for continuous EEG recording availability in the critical care setting.

Quantitative EEG analysis for automated detection of nonconvulsive seizures in intensive care units

Because of increased awareness of the high prevalence of nonconvulsive seizures in critically ill patients, use of continuous EEG (cEEG) monitoring is rapidly increasing in ICUs. However, cEEG monitoring is labor intensive, and manual review and interpretation of the EEG are impractical in most ICUs. Effective methods to assist in rapid and accurate detection of nonconvulsive seizures would greatly reduce the cost of cEEG monitoring and enhance the quality of patient care. In this study, we report a preliminary investigation of a novel ICU EEG analysis and seizure detection algorithm. Twenty-four prolonged cEEG recordings were included in this study. Seizure detection sensitivity and specificity were assessed for the new algorithm and for the two commercial seizure detection software systems. The new algorithm performed with a mean sensitivity of 90.4% and a mean false detection rate of 0.066/hour. The two commercial detection products performed with low sensitivities (12.9 and 10.1%) and false detection rates of 1.036/hour and 0.013/hour, respectively. These findings suggest that the novel algorithm has potential to be the basis of clinically useful software that can assist ICU staff in timely identification of nonconvulsive seizures. This study also suggests that currently available seizure detection software does not perform sufficiently in detection of nonconvulsive seizures in critically ill patients. This article is part of a Supplemental Special Issue entitled The Future of Automated Seizure Detection and Prediction.

A novel approach for computer assisted EEG monitoring in the adult ICU

OBJECTIVE

The implementation of a computer assisted system for real-time classification of the electroencephalogram (EEG) in critically ill patients.

METHODS

Eight quantitative features were extracted from the raw EEG and combined into a single classifier. The system was trained with 41 EEG recordings and subsequently evaluated using an additional 20 recordings. Through visual analysis, each recording was assigned to one of the following categories: normal, iso-electric, low voltage, burst suppression, slowing, and EEGs with generalized periodic discharges or seizure activity.

RESULTS

36 (88%) recordings from the training set and 17 (85%) recordings from the test set were classified correctly. A user interface was developed to present both trend-curves and a diagnostic output in text form. Implementation in a dedicated EEG monitor allowed real-time analysis in the intensive care unit (ICU) during pilot measurements in four patients.

CONCLUSIONS

We present the first results from a computer assisted EEG interpretation system, based on a combination of eight quantitative features. Our system provided an initial, reasonably accurate interpretation by non-experts of the most common EEG patterns observed in neurological patients in the adult ICU.

SIGNIFICANCE

Computer assisted EEG monitoring may improve early detection of seizure activity and ischemia in critically ill patients.

Therapeutic hypothermia and controlled normothermia in the intensive care unit: practical considerations, side effects, and cooling methods

BACKGROUND

Hypothermia is being used with increasing frequency to prevent or mitigate various types of neurologic injury. In addition, symptomatic fever control is becoming an increasingly accepted goal of therapy in patients with neurocritical illness. However, effectively controlling fever and inducing hypothermia poses special challenges to the intensive care unit team and others involved in the care of critically ill patients.

OBJECTIVE

To discuss practical aspects and pitfalls of therapeutic temperature management in critically ill patients, and to review the currently available cooling methods.

DESIGN

Review article.

INTERVENTIONS

None.

MAIN RESULTS

Cooling can be divided into three distinct phases: induction, maintenance, and rewarming. Each has its own risks and management problems. A number of cooling devices that have reached the market in recent years enable reliable maintenance and slow and controlled rewarming. In the induction phase, rapid cooling rates can be achieved by combining cold fluid infusion (1500-3000 mL 4 degrees C saline or Ringer's lactate) with an invasive or surface cooling device. Rapid induction decreases the risks and consequences of short-term side effects, such as shivering and metabolic disorders. Cardiovascular effects include bradycardia and a rise in blood pressure. Hypothermia's effect on myocardial contractility is variable (depending on heart rate and filling pressure); in most patients myocardial contractility will increase, although mild diastolic dysfunction can develop in some patients. A risk of clinically significant arrhythmias occurs only if core temperature decreases below 30 degrees C. The most important long-term side effects of hypothermia are infections (usually of the respiratory tract or wounds) and bedsores.

CONCLUSIONS

Temperature management and hypothermia induction are gaining importance in critical care medicine. Intensive care unit physicians, critical care nurses, and others (emergency physicians, neurologists, and cardiologists) should be familiar with the physiologic effects, current indications, techniques, complications and practical issues of temperature management, and induced hypothermia. In experienced hands the technique is safe and highly effective.

QEEG prognostic value in acute stroke

The objective of our study is to determine the predictive value of QEEG in patients suffering from an acute ischemic cerebral stroke. Twenty-eight patients were studied within the first 72 hours of clinical evolution of middle cerebral artery territory ischemic stroke. Thirty-seven QEEG recordings were obtained: 13 in the first 24 hours after cerebral stroke onset, 9 between 24-48 hours and 15 between 48-72 hours. Absolute Energies (AE) were the QEEG selected variables for statistical analysis: first, AE Z values were calculated using the Cuban QEEG norms, then the maximum and minimum AE Z values were selected within each frequency band and total power. The medians of the five neighboring Z values were also chosen. Regression models were estimated using the RANKIN scores as dependent variables and the selected QEEG variables as independent, then outcome predictions at hospital discharge and 3 months later were calculated. Percentages of concordance and errors between the estimated and real outcome scores were obtained. Alpha and theta AE were the best predictor for short-term outcome and delta AE for long-term outcome. We conclude that QEEG performed within the first 72 hours of ischemic stroke might be a powerful tool predicting short- and long-term outcome.

Assessment of hairline EEG as a screening tool for nonconvulsive status epilepticus

PURPOSE

Because of the high incidence of nonconvulsive status epilepticus (NCSE), the attraction of a "quick and easy" screening electroencephalogram (EEG) is obvious. Previous studies have shown utility of hairline EEG in diagnosing epilepsy. However, this technique has not been evaluated as a screening tool for NCSE. We wanted to provide proof of principle that a screening hairline EEG has sufficient sensitivity to use as a screening tool for diagnosing NCSE.

METHODS

A total of 120, 2- to 3-min EEG samples of normal and various abnormal digital EEG studies were reformatted in three six-channel montages (A, longitudinal bipolar; B, referential to ipsilateral ear; C, referential to contralateral ear) that mimicked a hairline recording and were interpreted by five neurophysiologists. The test data interpretation was compared with the original EEG interpretation.

RESULTS

Performance was best with montages A and B; 71% and 70.5% of the samples were interpreted correctly by using these montages. Only 65% of the samples were correctly interpreted by using montage C. With the best montage (A), the sensitivities ranged from 91% for normal EEG to 54% for periodic lateralized epileptiform discharges (PLEDs). The sensitivity for seizures was only 72%. Seizures were frequently misinterpreted as more benign patterns such as normal and diffuse slowing.

CONCLUSIONS

EEG data reformatted to resemble a hairline EEG had low sensitivity for detecting seizures. As a result, we do not recommend further pursuit of hairline EEG as a "quick and easy" screening tool for NCSE.

The interpretation of the EEG in stupor and coma

This review discusses a variety of causes of stupor and coma and associated electroencephalographic (EEG) findings. These include metabolic disturbances such as hepatic or renal dysfunction, which are often characterized by slowing of background rhythms and triphasic waves. Hypoxia and drug intoxications can produce a number of abnormal EEG patterns such as burst suppression, alpha coma, and spindle coma. Structural lesions, either supra- or infratentorial, are reviewed. EEGs in the former may show focal disturbances such as delta and theta activity, epileptiform abnormalities, and attenuation of faster frequencies. In infratentorial lesions, the EEG may appear normal, particularly with a pontine lesion. Some patients may be encephalopathic because of ongoing epileptic activity with minimal or no motor movements. This entity, nonconvulsive status epilepticus (NCSE), is difficult to diagnose in obtunded/comatose patients, and an EEG is required to verify the diagnosis and to monitor treatment. Several EEG patterns and their interpretation in suspected cases of NCSE such as periodic lateralized epileptiform discharges (PLEDs), bilateral independent periodic lateralized epileptiform discharges (BIPLEDs), generalized periodic epileptiform discharges (GPEDs), and triphasic waves are reviewed. Other entities discussed include the locked-in syndrome, neocortical death, persistent vegetative state, brainstem death, and brain death.

Continuous electroencephalogram monitoring in critically ill patients

The past few years have witnessed remarkable advances in continuous EEG monitoring (cEEG). The indications and applications for cEEG are broadening, including detection of nonconvulsive seizures, spell characterization, and prognostication. Seizures are common in the critically ill, are usually nonconvulsive, and can easily be missed without cEEG. Interpretation and clinical management of the complex periodic and rhythmic EEG patterns commonly identified in these patients require further study. With the use of quantitative analysis techniques, cEEG can detect cerebral ischemia very early, before permanent neuronal injury occurs. This article reviews the indications and recent advances in cEEG in critically ill patients. Continuous brain monitoring with cEEG is rapidly becoming the standard of care in critically ill patients with neurologic impairment.