Diagnostic utility of eight-channel EEG for detecting generalized or hemispheric seizures and rhythmic periodic patterns

Diagnostic accuracy of reduced electroencephalography montages for seizure detection: A frequentist and Bayesian meta-analysis

AIM

To evaluate the diagnostic accuracy of reduced montage electroencephalography (EEG) for seizure detection and provide evidence-based recommendations.

METHODS

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we conducted a diagnostic meta-analysis to assess the sensitivity and specificity of reduced EEG montages in detecting seizure activity. A hierarchical summary receiver operating characteristic curve (HSROC) model was used to estimate the area under the curve (AUC). Subgroup analyses were conducted to identify sources of heterogeneity. Bayesian estimates were used for validation.

RESULTS

Across 8 studies encompassing 3,458 reduced EEG montage samples, all studies used a reduced EEG montage with 7 to 10 electrodes. The pooled sensitivity was 0.75 (95 % CI: 0.68-0.80), and the pooled specificity was 0.97 (95 % CI: 0.95-0.98). The HSROC model had an AUC of 0.96 (95 % CI: 0.93-0.97). Variations in study results were attributed to factors such as the number of electrodes (pooled sensitivity of 0.66 for studies employing <8 leads and 0.77 for studies employing ≥8 leads) and montage design coverage (pooled sensitivity of 0.64 for studies employing subhairline montage and 0.77 for studies employing above-hairline montage). The Bayesian and frequentist findings agreed with each other and had a pooled sensitivity of 0.74 (95 % HPD: 0.65-0.83) and pooled specificity of 0.97 (95 % highest posterior density 0.95-0.98).

CONCLUSION

Reduced EEG montages with 8 or more electrodes are feasible for seizure detection, especially in emergency settings where rapid detection is crucial.

Point-of-Care Electroencephalography in Acute Neurological Care: A Narrative Review

Point-of-care electroencephalography (POC-EEG) systems are rapid-access, reduced-montage devices designed to address the limitations of conventional EEG (conv-EEG), enabling faster neurophysiological assessment in acute settings. This review evaluates their clinical impact, diagnostic performance, and feasibility in non-convulsive status epilepticus (NCSE), traumatic brain injury (TBI), stroke, and delirium. A comprehensive search of Medline, Scopus, and Embase identified 69 studies assessing 15 devices. In suspected NCSE, POC-EEG facilitates rapid seizure detection and prompt diagnosis, making it particularly effective in time-sensitive and resource-limited settings. Its after-hours availability and telemedicine integration ensure continuous coverage. AI-assisted tools enhance interpretability and accessibility, enabling use by non-experts. Despite variability in accuracy, it supports triaging, improving management, treatment decisions and outcomes while reducing hospital stays, transfers, and costs. In TBI, POC-EEG-derived quantitative EEG (qEEG) indices reliably detect structural lesions, support triage, and minimize unnecessary CT scans. They also help assess concussion severity and predict recovery. For strokes, POC-EEG aids triage by detecting large vessel occlusions (LVOs) with high feasibility in hospital and prehospital settings. In delirium, spectral analysis and AI-assisted models enhance diagnostic accuracy, broadening its clinical applications. Although POC-EEG is a promising screening tool, challenges remain in diagnostic variability, technical limitations, and AI optimization, requiring further research.

Accuracy of a Rapid-Response EEG's Automated Seizure-Burden Estimator: AccuRASE Study

BACKGROUND AND OBJECTIVES

The use of rapid response EEG (rr-EEG) has recently expanded in limited-resource settings and as a supplement to conventional EEG to rapidly detect and treat nonconvulsive status epilepticus. The study objective was to test the accuracy of an rr-EEG's automated seizure burden estimator (ASBE).

METHODS

This is a retrospective observational study using multiple blinded reviewers. All consecutive clinical rr-EEG procedures performed between November 2019 and February 2021 at Yale New Haven Hospital, one affiliated community hospital, and one affiliated inner-city regional hospital were included. Three reviewers blindly reviewed each EEG. The reference standard was 2/3 agreement. The co-primary outcome measures were the negative predictive value (NPV) of the ASBE for the detection of electrographic status epilepticus (ESE) or possible ESE (ESE/pESE) (to be used as a screening method to exclude ESE without the need for urgent expert review) and the positive predictive value (PPV, to be used for immediate treatment without requiring urgent expert review). These were assessed using a variety of seizure burden cutoffs determined by the algorithm (>1%, >10%, >20%, >50%, and >90%).

RESULTS

In the first 2 hours, a >10% burden cutoff detected 86% (95% CI 42%-100%) of studies with ESE alone and 88% (68%-97%) with ESE/pESE; this >10% cutoff had a NPV of 99% (97%-100%) for ESE and 98% (95%-100%) for ESE/pESE. The specificity at this threshold was 79% (73%-84%) for ESE and 84% (79%-89%) for ESE/pESE, but the PPV was low at 11% (4%-23%) for ESE and 39% (26%-53%) for ESE/pESE. A >90% burden cutoff was 97% (94%-99%) specific for detecting ESE (PPV 33% [7%-70%]) and 99% (97%-100%) specific for detecting ESE/pESE [PPV 78% (40%-97%)], although the sensitivity dropped significantly to 29% (13%-51%) for ESE/pESE and 43% (10%-82%) for ESE at the >90% threshold.

DISCUSSION

The ASBE has high specificity at >90% seizure burden threshold for detecting ESE and ESE/pESE, with good PPV for ESE/pESE, though with only low-to-moderate sensitivity; at this threshold, it can be used to help triage patients for immediate treatment/transfer, urgent expert review, and additional CEEG. A >10% threshold has a high sensitivity, detecting approximately 85% of patients with ESE; at this lower cutoff, it can be used as a screening tool to exclude ESE with >95% NPV.

CLASSIFICATION OF EVIDENCE

This study provides Class II evidence that ASBE software can reliably exclude ESE (98% negative predictive value using a <10% burden cutoff) without expert review in most patients requiring rapid response EEG.

Time is brain: detection of nonconvulsive seizures and status epilepticus during acute stroke evaluation using point-of-care electroencephalography

OBJECTIVES

Seizures are both a common mimic and a potential complication of acute stroke. Although EEG can be helpful to evaluate this differential diagnosis, conventional EEG infrastructure is resource-intensive and unable to provide timely monitoring to match the emergent context of a stroke code. We aimed to evaluate the real-world use and utility of a point-of-care EEG device as an adjunct to acute stroke evaluation.

MATERIALS AND METHODS

We performed a retrospective observational cohort study at a tertiary care community teaching hospital by identifying patients who underwent point-of-care EEG monitoring using Rapid Response EEG system (Ceribell Inc., Sunnyvale, CA) during stroke code evaluation of acute neurological deficits during the study period from January 1, 2020 to December 31, 2020. We assessed the frequency of seizures and highly epileptiform patterns among patients with either confirmed strokes or stroke mimics.

RESULTS

Point-of-care EEG monitoring was used in the wake of a stroke code in 70 patients. Of these, neuroimaging and clinical information resulted in a diagnosis of stroke in 38 patients (28 ischemic, 6 hemorrhagic, 4 transient ischemic attack; median NIHSS score of 6.5 [IQR 2.0-12.0]) and absence of any stroke in 32 patients. Point-of-care EEG detected seizures and highly epileptiform patterns in 6 (15.8 %) stroke patients and 11 (34.4 %) stroke-mimic patients, including 2 patients with persistent expressive aphasia due to repeated focal seizures.

CONCLUSIONS

Point-of-care EEG has utility for detecting nonconvulsive seizures in patients undergoing acute stroke evaluations.

Sensitivity of detecting interictal epileptiform activity using rapid reduced montage EEG

OBJECTIVE

Rapid EEG devices (REDs) have demonstrated substantial benefit regarding reduced time to performance of study and diagnosis in cases where urgent EEG is needed to evaluate patients for potentially revealing nonconvulsive status epilepticus and seizures. However, urgent EEG is also important in identifying cases regarding the need for initiation of antiseizure medication as well as triaging the use of continuous EEG monitoring. Some forms of REDs have a reduced montage (RRME) with electrode derivations that are one-half of standard recordings. This could impact spatial resolution and therefore potentially limit recovery of epileptiform abnormalities.

METHODS

In this study we evaluated the use of the Ceribell® rapid response EEG system and compared it to conventional video EEG (CvEEG). After applying inclusion and exclusion criteria, a total of 20 subjects were included in our analysis.

RESULTS

RRME was highly sensitive in detecting abundant and periodic discharges (p = 0.013) as well as discharges with a broad spatial distribution on CvEEG (p = 0.039). Sensitivity for detecting less prevalent discharges or those with more restricted spatial distribution was lower.

SIGNIFICANCE

Given the possibility of less frequent and more restricted epileptiform discharges eluding detection on RRME, we propose a protocol for the approach of using RRME and when to consider CvEEG when RRME is negative for epileptiform activity and highlight that urgent CvEEG may still be warranted following RRME.

Exploratory Study on the Clinical use of EEG for the People with Chronic Stroke and Their Correlation with the Neuropsychological Outcome

Objective. To measure the EEG signals of the people with chronic stroke in eyes-closed and eyes-open condition and study their relationship with the cognitive function and mental wellbeing. Methods. The investigators would conduct cognitive and mental wellbeing tests on recruited subjects. Their EEG signal was acquired by the 16-channel EEG system. The absolute power under different frequency bands and EEG indices (delta alpha ratio and pairwise derived brain symmetry index) in different eye conditions was calculated. Pearson's correlation was conducted to investigate the association between the clinical tests and the EEG index. Results. 32 subjects were recruited for the study. There was a significant correlation between the pairwise derived brain symmetry index (pdBSI) in eyes-open condition with the Stroop Test (p = .002), Paced Auditory Serial Addition Test-3 s (p = .008)/2 s (p = .002) and WHO-5 well-being scale (p = .023). Conclusions. There is a significant correlation between the brain symmetry index and the cognitive and wellbeing assessment. Brain symmetry index over the delta frequency has been found to be the most useful parameter relating to the clinical score.Significance:It is recommended to use EEG as an adjunctive neuropsychological assessment in clinics for people with chronic stroke, especially for clients who could not undertake conventional assessments (eg aphasia, attention problem).Highlights: There is a significant correlation between the EEG index and the clinical neuropsychological assessmentPairwise Derived Brain Symmetry index in delta frequency range correlated with most of the neuropsychological outcome.It is feasible for us to adopt EEG as an adjunctive assessment in clinical settings.

Landscape and future directions of machine learning applications in closed-loop brain stimulation

Brain stimulation (BStim) encompasses multiple modalities (e.g., deep brain stimulation, responsive neurostimulation) that utilize electrodes implanted in deep brain structures to treat neurological disorders. Currently, BStim is primarily used to treat movement disorders such as Parkinson's, though indications are expanding to include neuropsychiatric disorders like depression and schizophrenia. Traditional BStim systems are "open-loop" and deliver constant electrical stimulation based on manually-determined parameters. Advancements in BStim have enabled development of "closed-loop" systems that analyze neural biomarkers (e.g., local field potentials in the sub-thalamic nucleus) and adjust electrical modulation in a dynamic, patient-specific, and energy efficient manner. These closed-loop systems enable real-time, context-specific stimulation adjustment to reduce symptom burden. Machine learning (ML) has emerged as a vital component in designing these closed-loop systems as ML models can predict / identify presence of disease symptoms based on neural activity and adaptively learn to modulate stimulation. We queried the US National Library of Medicine PubMed database to understand the role of ML in developing closed-loop BStim systems to treat epilepsy, movement disorders, and neuropsychiatric disorders. Both neural and non-neural network ML algorithms have successfully been leveraged to create closed-loop systems that perform comparably to open-loop systems. For disorders in which the underlying neural pathophysiology is relatively well understood (e.g., Parkinson's, essential tremor), most work has involved refining ML models that can classify neural signals as aberrant or normal. The same is seen for epilepsy, where most current research has focused on identifying optimal ML model design and integrating closed-loop systems into existing devices. For neuropsychiatric disorders, where the underlying pathologic neural circuitry is still being investigated, research is focused on identifying biomarkers (e.g., local field potentials from brain nuclei) that ML models can use to identify onset of symptoms and stratify severity of disease.

Risk of Developing Seizures in Children With Abnormal EEG Findings During Polysomnography

BACKGROUND

Polysomnography (PSG) utilizes abbreviated electroencephalogram (EEG) to stage sleep. The aim of this study was to determine whether epileptiform abnormalities on this limited EEG coverage correlated with abnormalities on routine EEG (rEEG) and an increased risk for seizures in children without a prior diagnosis of epilepsy.

METHODS

A six-year retrospective chart review was performed assessing children with abnormalities on EEG during PSG. Children who underwent subsequent rEEG were included; children with a prior diagnosis of seizures were excluded. The main outcome measures were rEEG results and subsequent diagnosis of epilepsy.

RESULTS

A total of 67 children met inclusion criteria. Average age was six years, and 43 (64%) were male. rEEG was normal in 16 (24%). Epileptiform abnormalities were focal in 36 (54%), generalized in eight (12%), and mixed in five (8%). An additional two (3%) had slow background rhythm without epileptiform discharges. Thirty-one patients had neurology clinic follow-up with an average duration of 31 months (range 4 to 65 months). Of these, nine (29%) developed seizures, including all three with generalized epileptiform discharges, four of 19 (21%) with focal epileptiform discharges, and two of five (40%) with mixed epileptiform discharges or background slowing. None of the four patients with a normal rEEG had seizures. Eight of the nine patients with seizures were treated with antiepileptic drugs.

CONCLUSIONS

Children with no history of seizures found to have abnormal EEG during PSG are likely to have an abnormal rEEG. Additionally, they have an increased risk for developing seizures.

Quantitative detection of seizures with minimal-density EEG montage using phase synchrony and cross-channel coherence amplitude in critical care

Seizures frequently occur in paediatric emergency and critical care, with up to 74% being sub-clinical seizures making detection difficult. Delays in seizure detection and treatment worsen the neurological outcome of critically-ill patients. Gold-standard seizure detections using multi-channels electroencephalograms (EEG) require trained clinical physiologists to apply scalp electrodes and highly specialised neurologists to interpret and identify seizures. In this study, we extracted phase synchrony and cross-channel coherence amplitude across 4 and 8 pre-selected scalp EEG signals. Binary classification is used to determine whether the signal segment is seizure or non-seizure, and the predictions were compared against the gold-standard seizure onset markings. The application of the algorithm on a cohort of forty routinely collected EEGs from paediatric patients showed an average accuracy of 77.2 % and 76.5% using 4 and 8 channels, respectively. Clinical Relevance- This work demonstrates the feasibility of seizure detection with pre-defined 4 and 8 EEG electrodes with an average accuracy of 77%. This means for the first time seizure detection is possible using an EEG montage that can be applied readily at the bedside independent of expert input.

Comparing Seizures Captured by Rapid Response EEG and Conventional EEG Recordings in a Multicenter Clinical Study

Objective

A recent multicenter prospective study (DECIDE trial) examined the use of Ceribell Rapid Response EEG (Rapid-EEG) in the emergent evaluation and management of critically ill patients suspected to have non-convulsive seizures. We present a detailed, patient-level examination of seizures detected either on initial Rapid-EEG or subsequent conventional EEG within 24 h to investigate whether seizures were missed on Rapid-EEG due to the exclusion of midline/parasagittal coverage.

Methods

We identified from 164 patients studied in the DECIDE trial those who had seizures detected on Rapid-EEG but not conventional EEG (n = 6), conventional EEG but not Rapid-EEG (n = 4), or both Rapid-EEG and conventional EEG (n = 9). We examined the electrographic characteristics of ictal and interictal findings on both devices, especially their detection in lateral or midline/parasagittal chains, and patient clinical histories to identify contributors toward discordant seizure detection.

Results

Seizures detected on both EEG systems had similar electrographic appearance and laterality. Seizures detected only on conventional EEG (within 24 h following Rapid-EEG) were visible in the temporal chains, and external clinical factors (e.g., treatment with anti-seizure medications, sedation, and duration of recordings) explained the delayed presentation of seizures. Patients with seizures detected only by Rapid-EEG were treated with anti-seizure medications, and subsequent conventional EEG detected interictal highly epileptiform patterns with similar laterality.

Conclusions

Our case series demonstrates that electrographic data obtained from initial Rapid-EEG and subsequent conventional EEG monitoring are largely concordant relative to morphology and laterality. These findings are valuable to inform future investigation of abbreviated EEG systems to optimize management of suspected non-convulsive seizures and status epilepticus. Future, larger studies could further investigate the value of Rapid-EEG findings for forecasting and predicting seizures in long-term EEG recordings.

Improved access to rapid electroencephalography at a community hospital reduces inter-hospital transfers for suspected non-convulsive seizures

OBJECTIVE

Patients with suspected non-convulsive seizures are optimally evaluated with EEG. However, limited EEG infrastructure at community hospitals often necessitates transfer for long-term EEG monitoring (LTM). Novel point-of-care EEG systems could expedite management of nonconvulsive seizures and reduce unnecessary transfers. We aimed to describe the impact of rapid access to EEG using a novel EEG device with remote expert interpretation (tele-EEG) on rates of transfer for LTM.

METHODS

We retrospectively identified a cohort of patients who underwent Rapid-EEG (Ceribell Inc., Mountain View, CA) monitoring as part of a new standard-of-care at a community hospital. Rapid-EEGs were initially reviewed on-site by a community hospital neurologist before transitioning to tele-EEG review by epileptologists at an affiliated academic hospital. We compared the rate of transfer for LTM after Rapid-EEG/tele-EEG implementation to the expected rate if rapid access to EEG was unavailable.

RESULTS

Seventy-four patients underwent a total of 118 Rapid-EEG studies (10 with seizure, 18 with highly epileptiform patterns, 90 with slow/normal activity). Eighty-one studies (69%), including 9 of 10 studies that detected seizures, occurred after-hours when EEG was previously unavailable. Based on historical practice patterns, we estimated that Rapid-EEG potentially obviated transfer for LTM in 31 of 33 patients (94%); both completed transfers occurred before the transition to tele-EEG review.

SIGNIFICANCE

Rapid access to EEG led to the detection of seizures that would otherwise have been missed and reduced inter-hospital transfers for LTM. We estimate that the reduction in inter-hospital transportation costs alone would be in excess of $39,000 ($1,274 per patient). Point-of-care EEG systems may support a hub-and-spoke model for managing non-convulsive seizures (similar to that utilized in this study and analogous to existing acute stroke infrastructures), with increased EEG capacity at community hospitals and tele-EEG interpretation by specialists at academic hospitals that can accept transfers for LTM.

Effects of a reduction of the number of electrodes in the EEG montage on the number of identified seizure patterns

Continuous EEG monitoring (cEEG) is frequently used in neurocritical care. The detection of seizures is one of the main objectives. The placement of the EEG electrodes is time consuming, therefore a reduced montage might lead to an increased availability in the ICU setting. It is unknown whether such a reduction of electrodes reduces the number of seizure patterns that are detected. A total of 95 seizure and 95 control EEG sequences from a pediatric epilepsy monitoring unit (EMU) were anonymized and reduced to an eight-lead montage. Two experts evaluated the recordings and the seizure detection rates using the reduced and the full montage were compared. Sensitivity and specificity for the seizure detection were calculated using the original EMU findings as gold standard. The sensitivity to detect seizures was 0.65 for the reduced montage compared to 0.76 for the full montage (p = 0.031). The specificities (0.97 and 0.96) were comparable (p = 1). A total of 4/9 (44%) of the generalized, 12/44 (27%) of the frontal, 6/14 (43%) of the central, 0/1 (0%) of the occipital, 6/20 (30%) of the temporal, and 5/7 (71%) of the parietal seizure patterns were not detected using the reduced montage. The median time difference between the onset of the seizure pattern in the full and reduced montage was 0.026s (IQR 5.651s). In this study the reduction of the EEG montage from 21 to eight electrodes reduced the sensitivity to detect seizure patterns from 0.76 to 0.65. The specificity remained virtually unchanged.

Resident Use of EEG Cap System to Rule Out Nonconvulsive Status Epilepticus

BACKGROUND

Nonconvulsive status epilepticus (NCSE) requires an EEG for diagnosis and in many centers access may be limited. The authors aimed to test whether neurology residents can be trained to use and interpret full-montage EEGs using an EEG cap electrode system to detect NCSE while on-call.

METHODS

Neurology residents were trained to interpret EEG recordings using the American Clinical Neurophysiology Society critical care EEG terminology. Residents who achieved a score of 70% or higher in the American Clinical Neurophysiology Society certification test and attended a training session were eligible to use the EEG cap on-call with patients suspected of having NCSE. Residents' experience and interpretation of observed EEG patterns were evaluated using a questionnaire. Each EEG recording was independently reviewed by three epilepsy specialists to determine the interpretability of each study and whether the residents correctly identified the EEG patterns.

RESULTS

Sixteen residents undertook the training and 12 (75%) achieved a score of 70% or higher on the certification test. Seven of these residents performed 14 EEG cap studies between August 2017 and May 2018. The percent agreement between residents and electroencephalographers was 78.6% for EEG interpretability and 57.1% for description of EEG pattern. Residents did not miss any malignant patterns concerning for NCSE, which accounted for 1 of 14 EEGs but "overcalled" patterns as malignant in 3 of 14 recordings.

CONCLUSIONS

This study suggests that neurology residents can be taught to perform and interpret EEGs using a cap system to monitor for NCSE. Additional training will help improve EEG interpretation and sensitivity.

The Utility of Infectious and Neurodiagnostic Testing in Children With Complex Febrile Seizures Requiring Mechanical Ventilation

A retrospective cohort analysis was performed on 79 consecutive patients between 6 months and 5 years admitted to a tertiary hospital with a diagnosis of complex febrile seizures requiring mechanical ventilation from 2011 to 2017 to determine the utility of infectious and neurologic diagnostics. Intubation was used as a proxy for severity of illness. The overall intensive care unit stay was short (95% intubated <24 hours, 88% admitted <3 days). No life-threatening infections were identified, and none required surgical interventions. Electroencephalogram (EEG) was obtained on 43%, 26% of which were abnormal. Sixty-six percent of patients were discharged on rescue benzodiazepine and 20% with maintenance antiseizure medications. Duration of follow-up averaged 4 years (range 1 month to 9 years); 8 patients (10%) were subsequently diagnosed with epilepsy. Our findings suggest that extensive diagnostic evaluations may not be necessary for children with complex febrile seizures requiring mechanical ventilation although the role of EEG is less understood.

Evaluating the Clinical Impact of Rapid Response Electroencephalography: The DECIDE Multicenter Prospective Observational Clinical Study

OBJECTIVES

To measure the diagnostic accuracy, timeliness, and ease of use of Ceribell rapid response electroencephalography. We assessed physicians' diagnostic assessments and treatment plans before and after rapid response electroencephalography assessment. Primary outcomes were changes in physicians' diagnostic and therapeutic decision making and their confidence in these decisions based on the use of the rapid response electroencephalography system. Secondary outcomes were time to electroencephalography, setup time, ease of use, and quality of electroencephalography data.

DESIGN

Prospective multicenter nonrandomized observational study.

SETTING

ICUs in five academic hospitals in the United States.

SUBJECTS

Patients with encephalopathy suspected of having nonconvulsive seizures and physicians evaluating these patients.

INTERVENTIONS

Physician bedside assessment of sonified electroencephalography (30 s from each hemisphere) and visual electroencephalography (60 s) using rapid response electroencephalography.

MEASUREMENTS AND MAIN RESULTS

Physicians (29 fellows or residents, eight attending neurologists) evaluated 181 ICU patients; complete clinical and electroencephalography data were available in 164 patients (average 58.6 ± 18.7 yr old, 45% females). Relying on rapid response electroencephalography information at the bedside improved the sensitivity (95% CI) of physicians' seizure diagnosis from 77.8% (40.0%, 97.2%) to 100% (66.4%, 100%) and the specificity (95% CI) of their diagnosis from 63.9% (55.8%, 71.4%) to 89% (83.0%, 93.5%). Physicians' confidence in their own diagnosis and treatment plan were also improved. Time to electroencephalography (median [interquartile range]) was 5 minutes (4-10 min) with rapid response electroencephalography while the conventional electroencephalography was delayed by several hours (median [interquartile range] delay = 239 minutes [134-471 min] [p < 0.0001 using Wilcoxon signed rank test]). The device was rated as easy to use (mean ± SD: 4.7 ± 0.6 [1 = difficult, 5 = easy]) and was without serious adverse effects.

CONCLUSIONS

Rapid response electroencephalography enabled timely and more accurate assessment of patients in the critical care setting. The use of rapid response electroencephalography may be clinically beneficial in the assessment of patients with high suspicion for nonconvulsive seizures and status epilepticus.

Virtual EEG-electrodes: Convolutional neural networks as a method for upsampling or restoring channels

BACKGROUND

In clinical practice, EEGs are assessed visually. For practical reasons, recordings often need to be performed with a reduced number of electrodes and artifacts make assessment difficult. To circumvent these obstacles, different interpolation techniques can be utilized. These techniques usually perform better for higher electrode densities and values interpolated at areas far from electrodes can be unreliable. Using a method that learns the statistical distribution of the cortical electrical fields and predicts values may yield better results.

NEW METHOD

Generative networks based on convolutional layers were trained to upsample from 4 or 14 channels or to dynamically restore single missing channels to recreate 21-channel EEGs. 5,144 h of data from 1,385 subjects of the Temple University Hospital EEG database were used for training and evaluating the networks.

COMPARISON WITH EXISTING METHOD

The results were compared to spherical spline interpolation. Several statistical measures were used as well as a visual evaluation by board certified clinical neurophysiologists. Overall, the generative networks performed significantly better. There was no difference between real and network generated data in the number of examples assessed as artificial by experienced EEG interpreters whereas for data generated by interpolation, the number was significantly higher. In addition, network performance improved with increasing number of included subjects, with the greatest effect seen in the range 5-100 subjects.

CONCLUSIONS

Using neural networks to restore or upsample EEG signals is a viable alternative to spherical spline interpolation.

Ceribell EEG shortens seizure diagnosis and workforce time and is useful for COVID isolation

OBJECTIVE

To determine whether the portable Ceribell^® electroencephalograph (EEG) (Mountain View, CA) used for suspected status epilepticus (SE) can reduce time to diagnosis and on-call workforce demands and whether it can be applied to patients in respiratory isolation.

METHODS

A multidisciplinary team developed a protocol for the use of the Ceribell EEG. The staff deploying the device, the attending physician, and the interpreting neurologist completed evaluation tools for each patient. Data maintained for quality and resource planning of 18-channel electroencephalography ordered for suspected SE were used as controls. Times to diagnosis were compared by application of Welch-Satterthwaite tests and workforce call-in demands by Fisher's exact t test. We evaluated qualitative data related to the use of the EEG in COVID-19 isolation rooms and on its technical aspects and acceptance by staff members.

RESULTS

The Ceribell EEG reduced diagnosis time (P = .0000006) and on-call workforce demand (P = .02). The device can be used at any time of day in any hospital care area and has advantages in respiratory isolation rooms.

SIGNIFICANCE

Compared with a standard 18-channel EEG, the Ceribell device allowed earlier diagnosis of SE and non-SE conditions and reduced workforce demands. Due to the ease of its use and its simple components, which can be readily disinfected, it is advantageous for COVID-19 patients in isolation.

Structure and Outcomes of Educational Programs for Training Non-electroencephalographers in Performing and Screening Adult EEG: A Systematic Review

OBJECTIVE

To qualitatively and quantitatively summarize curricula, teaching methods, and effectiveness of educational programs for training bedside care providers (non-experts) in the performance and screening of adult electroencephalography (EEG) for nonconvulsive seizures and other patterns.

METHODS

PRISMA methodological standards were followed. MEDLINE, EMBASE, Cochrane, CINAHL, WOS, Scopus, and MedEdPORTAL databases were searched from inception until February 26, 2020 with no restrictions. Abstract and full-text review was completed in duplicate. Studies were included if they were original research; involved non-experts performing, troubleshooting, or screening adult EEG; and provided qualitative descriptions of curricula and teaching methods and/or quantitative assessment of non-experts (vs gold standard EEG performance by neurodiagnostic technologists or interpretation by neurophysiologists). Data were extracted in duplicate. A content analysis and a meta-narrative review were performed.

RESULTS

Of 2430 abstracts, 35 studies were included. Sensitivity and specificity of seizure identification varied from 38 to 100% and 65 to 100% for raw EEG; 40 to 93% and 38 to 95% for quantitative EEG, and 95 to 100% and 65 to 85% for sonified EEG, respectively. Non-expert performance of EEG resulted in statistically significant reduced delay (86 min, p < 0.0001; 196 min, p < 0.0001; 667 min, p < 0.005) in EEG completion and changes in management in approximately 40% of patients. Non-experts who were trained included physicians, nurses, neurodiagnostic technicians, and medical students. Numerous teaching methods were utilized and often combined, with instructional and hands-on training being most common.

CONCLUSIONS

Several different bedside providers can be educated to perform and screen adult EEG, particularly for the purpose of diagnosing nonconvulsive seizures. While further rigorous research is warranted, this review demonstrates several potential bridges by which EEG may be integrated into the care of critically ill patients.

Modeling the economic value of Ceribell Rapid Response EEG in the inpatient hospital setting

AIMS

Potentially life-threatening diagnosis of non-convulsive status epilepticus (NCSE) can only be confirmed with electroencephalography (EEG). When access to EEG is limited, physicians may empirically treat, risking unnecessary sedation and intubation, or not treat, increasing risk of refractory seizures. Either may prolong hospital length of stay (LOS). The current study aimed to examine the effect of a new EEG system (Ceribell Rapid Response EEG, Rapid-EEG) on hospital costs by enabling easy access to EEG and expedited seizure diagnosis and treatment.

MATERIALS AND METHODS

We built a two-armed decision-analytic cost-benefit model comparing Rapid-EEG with clinical suspicion alone for NCSE. Diagnostic parameters were informed by a multicenter clinical trial (DECIDE, NCT03534258), while LOS and cost parameters were from public US inpatient data, published literature, and Center for Medicare and Medicaid Services fee schedules. We calculated reference case estimates from mean values, while uncertainty was assessed using 95% prediction intervals (PI) generated by probabilistic sensitivity analysis (PSA) and ANCOVA sum of squares. All costs were indexed to 2019 US$.

RESULTS

Each use case of Rapid-EEG saved $3,971 to $17,290 as it led to reduction in the hospital LOS by 1.2 days (6.1 vs. 7.4 days) and ICU LOS by 0.4 days (1.5 vs. 1.9 days). Using PSA, Rapid-EEG saving was $5,633 per use case (95% PI: $($4,649 to $6,617), as it led to diminished hospital LOS by 1.1 days (95% PI: 0.9-1.4 days) and reduced ICU LOS by 0.5 days (95% PI: 0.4-0.6 days). Cost-savings were demonstrated in 75% of replications. Sixty-four percent of variance in total costs was attributable to LOS for persons incorrectly diagnosed with seizures.

LIMITATIONS

Results were obtained from the analysis of existing data and not a prospective outcome trial.

CONCLUSIONS

Rapid-EEG alters the treatment course for patients with suspected seizures and will result in cost savings per patient.

Monitoring the Burden of Seizures and Highly Epileptiform Patterns in Critical Care with a Novel Machine Learning Method

INTRODUCTION

Current electroencephalography (EEG) practice relies on interpretation by expert neurologists, which introduces diagnostic and therapeutic delays that can impact patients' clinical outcomes. As EEG practice expands, these experts are becoming increasingly limited resources. A highly sensitive and specific automated seizure detection system would streamline practice and expedite appropriate management for patients with possible nonconvulsive seizures. We aimed to test the performance of a recently FDA-cleared machine learning method (Claritγ, Ceribell Inc.) that measures the burden of seizure activity in real time and generates bedside alerts for possible status epilepticus (SE).

METHODS

We retrospectively identified adult patients (n = 353) who underwent evaluation of possible seizures with Rapid Response EEG system (Rapid-EEG, Ceribell Inc.). Automated detection of seizure activity and seizure burden throughout a recording (calculated as the percentage of ten-second epochs with seizure activity in any 5-min EEG segment) was performed with Claritγ, and various thresholds of seizure burden were tested (≥ 10% indicating ≥ 30 s of seizure activity in the last 5 min, ≥ 50% indicating ≥ 2.5 min of seizure activity, and ≥ 90% indicating ≥ 4.5 min of seizure activity and triggering a SE alert). The sensitivity and specificity of Claritγ's real-time seizure burden measurements and SE alerts were compared to the majority consensus of at least two expert neurologists.

RESULTS

Majority consensus of neurologists labeled the 353 EEGs as normal or slow activity (n = 249), highly epileptiform patterns (HEP, n = 87), or seizures [n = 17, nine longer than 5 min (e.g., SE), and eight shorter than 5 min]. The algorithm generated a SE alert (≥ 90% seizure burden) with 100% sensitivity and 93% specificity. The sensitivity and specificity of various thresholds for seizure burden during EEG recordings for detecting patients with seizures were 100% and 82% for ≥ 50% seizure burden and 88% and 60% for ≥ 10% seizure burden. Of the 179 EEG recordings in which the algorithm detected no seizures, seizures were identified by the expert reviewers in only two cases, indicating a negative predictive value of 99%.

DISCUSSION

Claritγ detected SE events with high sensitivity and specificity, and it demonstrated a high negative predictive value for distinguishing nonepileptiform activity from seizure and highly epileptiform activity.

CONCLUSIONS

Ruling out seizures accurately in a large proportion of cases can help prevent unnecessary or aggressive over-treatment in critical care settings, where empiric treatment with antiseizure medications is currently prevalent. Claritγ's high sensitivity for SE and high negative predictive value for cases without epileptiform activity make it a useful tool for triaging treatment and the need for urgent neurological consultation.

Diagnostic Value of Electroencephalography with Ten Electrodes in Critically Ill Patients

BACKGROUND

In critical care settings, electroencephalography (EEG) with reduced number of electrodes (reduced montage EEG, rm-EEG) might be a timely alternative to the conventional full montage EEG (fm-EEG). However, past studies have reported variable accuracies for detecting seizures using rm-EEG. We hypothesized that the past studies did not distinguish between differences in sensitivity from differences in classification of EEG patterns by different readers. The goal of the present study was to revisit the diagnostic value of rm-EEG when confounding issues are accounted for.

METHODS

We retrospectively collected 212 adult EEGs recorded at Massachusetts General Hospital and reviewed by two epileptologists with access to clinical, trending, and video information. In Phase I of the study, we re-configured the first 4 h of the EEGs in lateral circumferential montage with ten electrodes and asked new readers to interpret the EEGs without access to any other ancillary information. We compared their rating to the reading of hospital clinicians with access to ancillary information. In Phase II, we measured the accuracy of the same raters reading representative samples of the discordant EEGs in full and reduced configurations presented randomly by comparing their performance to majority consensus as the gold standard.

RESULTS

Of the 95 EEGs without seizures in the selected fm-EEG, readers of rm-EEG identified 92 cases (97%) as having no seizure activity. Of 117 EEGs with "seizures" identified in the selected fm-EEG, none of the cases was labeled as normal on rm-EEG. Readers of rm-EEG reported pathological activity in 100% of cases, but labeled them as seizures (N = 77), rhythmic or periodic patterns (N = 24), epileptiform spikes (N = 7), or burst suppression (N = 6). When the same raters read representative epochs of the discordant EEG cases (N = 43) in both fm-EEG and rm-EEG configurations, we found high concordance (95%) and intra-rater agreement (93%) between fm-EEG and rm-EEG diagnoses.

CONCLUSIONS

Reduced EEG with ten electrodes in circumferential configuration preserves key features of the traditional EEG system. Discrepancies between rm-EEG and fm-EEG as reported in some of the past studies can be in part due to methodological factors such as choice of gold standard diagnosis, asymmetric access to ancillary clinical information, and inter-rater variability rather than detection failure of rm-EEG as a result of electrode reduction per se.

How soon should urgent EEG be performed following a first epileptic seizure?

PURPOSE

Patients with a first unprovoked epileptic seizure are often seen in emergency services. Electroencephalography (EEG) is indicated for diagnosing epilepsy, but the optimal time to perform this test has not been defined. This study aimed to determine the time interval following a seizure within which EEG has the greatest diagnostic yield.

METHODS

We conducted a retrospective study of all adult patients with a first unprovoked seizure who had undergone emergency EEG (July 2014-December 2019). Data collection included demographics, seizure type, time interval to EEG study, EEG pattern identified, and the prescription after emergency assessment. An optimal cut-off point for time to EEG was obtained, and an adjusted regression model was performed to establish associations with the presence of epileptiform abnormalities.

RESULTS

A total of 170 patients were included (mean age: 50.7 years, 40.6% women). Epileptiform discharges were identified in 34.1% of recordings, nonepileptiform abnormalities in 46.5%, and normal findings in 19.4%. A lower latency from seizure to EEG was associated with a higher probability of finding epileptiform discharges (median: 12.7 in the epileptiform EEGs vs. 20 h in the nonepileptiform EEGs, p < 0.001). The time interval associated with the highest probability of detecting an epileptiform EEG pattern was within the first 16 h after seizure onset: 52.1% of recordings performed before the 16-h cut-off showed these abnormal patterns compared with 20.2% performed after (p < 0.001). These findings were not related to the presence of an epileptogenic lesion in neuroimaging or to other clinical variables. The finding of epileptiform abnormalities was followed by a greater prescription of antiseizure drugs (96.4% vs. 66% in nonepileptiform patterns, p < 0.001).

CONCLUSION

The diagnostic yield of EEG following a first unprovoked epileptic seizure is highest when this test is performed within the first 16 h after onset of the event.

Electroencephalographic findings in COVID-19 patients: A systematic review

Background

Growing evidence of neurologic involvement seen in COVID-19 infection necessitates the pooling of neurodiagnostic findings like electroencephalography (EEG) that may guide clinical management. The objective of this study was to review the EEG findings in patients diagnosed with COVID-19 infection through a systematic review of published articles.

Methods

We systematically searched until July 25, 2020 for published articles that reported on descriptive EEG findings in patients diagnosed with COVID-19 in PUBMED by Medline, EMBASE, and CENTRAL by the Cochrane Library.

Results

From a total of 94 identified records, 29 relevant articles were included in this review. A total of 177 patients with COVID-19 with descriptive EEG reports were analyzed. The most common indication for EEG was unexplained altered mental status. Disturbances of background activity such as generalized and focal slowing were seen as well epileptiform abnormalities and rhythmic or periodic discharges. There were no consistent EEG findings specific to COVID-19 infection.

Conclusion

The EEG findings in COVID-19 appear to be non-specific. Further research on the relationship of the EEG findings to the clinical state and short- or long-term prognosis of COVID-19 patients may be conducted to help clinicians discern which patients would necessitate an EEG procedure and would eventually require treatment.

Validation of an algorithm of time-dependent electro-clinical risk stratification for electrographic seizures (TERSE) in critically ill patients

OBJECTIVE

The clinical implementation of continuous electroencephalography (CEEG) monitoring in critically ill patients is hampered by the substantial burden of work that it entails for clinical neurophysiologists. Solutions that might reduce this burden, including by shortening the duration of EEG to be recorded, would help its widespread adoption. Our aim was to validate a recently described algorithm of time-dependent electro-clinical risk stratification for electrographic seizure (ESz) (TERSE) based on simple clinical and EEG features.

METHODS

We retrospectively reviewed the medical records and EEG recordings of consecutive patients undergoing CEEG between October 1, 2015 and September, 30 2016 and assessed the sensitivity of TERSE for seizure detection, as well as the reduction in EEG time needed to be reviewed.

RESULTS

In a cohort of 407 patients and compared to full CEEG review, the model allowed the detection of 95% of patients with ESz and 97% of those with electrographic status epilepticus. The amount of CEEG to be recorded to detect ESz was reduced by two-thirds, compared to the duration of CEEG taht was actually recorded.

CONCLUSIONS

TERSE allowed accurate time-dependent ESz risk stratification with a high sensitivity for ESz detection, which could substantially reduce the amount of CEEG to be recorded and reviewed, if applied prospectively in clinical practice.

SIGNIFICANCE

Time-dependent electro-clinical risk stratification, such as TERSE, could allow more efficient practice of CEEG and its more widespread adoption. Future studies should aim to improve risk stratification in the subgroup of patients with acute brain injury and absence of clinical seizures.

Practical Considerations When Performing Neurodiagnostic Studies on Patients with COVID-19 and Other Highly Virulent Diseases

The coronavirus disease 2019, SARS-COV-2 (the cause of COVID-19), has led to a worldwide shortage of personal protective equipment (PPE) and an increased stress on hospital resources, which has resulted in a spike in the anxiety of the frontline healthcare workers. News reports and information about the virus are rapidly changing. We present a case of a patient with COVID-19 who had a seizure-like spell for which an EEG was performed. In early to mid-March, there were no clear guidelines or recommendations available from neurodiagnostic-related organizations or hospitals on how to adapt procedure workflow to those with COVID-19. When caring for COVID-19 patients, as when caring for any patient with an infectious disease, it is hospital protocol to follow contact, droplet/airborne precautions by wearing appropriate PPE. However, because we knew very little about the coronavirus, this case was different. In this article, we discuss our experience with our EEG workflow and concerns for staff exposure. We then discuss our adaptations and modifications to our standard procedures and protocols. A time analysis comparing our standard EEG protocol with our modified COVID-19 protocol revealed a significant decrease in technologist exposure time (99 minutes versus 51 minutes), which theoretically would reduce the chance of virus transmission to our technologist. At this critical moment in time, we hope such modifications will allow us to continue delivering high quality patient care while optimizing resource utilization and above all keeping our technologists safe.

Infection Prevention: 2020 Review and Update for Neurodiagnostic Technologists

Since 1995, ASET has published recommendations for infection prevention. With the aim of keeping our readers current with updates in infection prevention initiatives, this article reviews ASET's past publications by Altman 1995, Altman 2000, Scott 2013, and Sullivan & Altman 2008, and incorporates new information from published scientific literature, online resources, print publications, national and international guidelines, OSHA and other regulatory agencies. Knowledge of current infection control practices and recommendations is essential for every Neurodiagnostic technologist, whether working in a hospital, an ambulatory setting, intensive care unit or in the operating room. All technologists who have direct patient contact are responsible for ensuring use of best practices to prevent the spread of infection.

Time Is Brain: The Use of EEG Electrode Caps to Rapidly Diagnose Nonconvulsive Status Epilepticus

OBJECTIVE

To perform a feasibility pilot study comparing the usefulness of EEG electrode cap versus standard scalp EEG for acquiring emergent EEGs in emergency department, inpatient, and intensive care unit patients.

BACKGROUND

Nonconvulsive status epilepticus (NCSE) is a neurological emergency diagnosed exclusively by EEG. Nonconvulsive status epilepticus becomes more resistant to treatment 1 hour after continued seizure activity. EEG technologists are alerted "stat" when there is immediate need for an EEG during oncall hours, yet delays are inevitable. Alternatively, EEG caps can be quickly placed by in-house residents at bedside for assessment.

DESIGN/METHODS

EEG caps were compared with standard-of-care "stat" EEGs for 20 patients with suspected NCSE. After the order for a stat EEG was placed, neurology residents were simultaneously alerted and placed an EEG cap prior to the arrival of the on-call out-of-hospital technologist. Both EEG cap recordings and standard EEG recordings were visually reviewed at 10 and 20 minutes in a blinded manner by two electroencephalographers. The timing, accuracy of interpretation, and diagnosis between the two techniques were then compared.

RESULTS

Of the 20 adult patients, 70% (14 of 20) of EEG cap recordings were interpretable, whereas 95% (19 of 20) standard EEGs were interpretable; three had findings consistent with NCSE on both the EEG cap and standard EEG recordings. In the time analysis, 16 patients were included. EEG cap placement was significantly more time efficient than an EEG performed by technologist using the usual "stat" EEG protocol, with the median EEG cap electrode placement occurring 86 minutes faster than standard EEG (22.5 minutes vs. 104.5 minutes; P < 0.0001; n = 16).

CONCLUSIONS

New rapid EEG recording using improved EEG caps may allow for rapid diagnosis and clinical decision making in suspected NCSE.

Rapid Bedside Evaluation of Seizures in the ICU by Listening to the Sound of Brainwaves: A Prospective Observational Clinical Trial of Ceribell's Brain Stethoscope Function

BACKGROUND

Patients suffering from non-convulsive seizures experience delays in diagnosis and treatment due to limitations in acquiring and interpreting electroencephalography (EEG) data. The Ceribell EEG System offers rapid EEG acquisition and conversion of EEG signals to sound (sonification) using a proprietary algorithm. This study was designed to test the performance of this EEG system in an intensive care unit (ICU) setting and measure its impact on clinician treatment decision.

METHODS

Encephalopathic ICU patients at Stanford University Hospital were enrolled if clinical suspicion for seizures warranted EEG monitoring. Treating physicians rated suspicion for seizure and decided if the patient needed antiepileptic drug (AED) treatment at the time of bedside evaluation. After listening to 30 s of EEG from each hemisphere in each patient, they reevaluated their suspicion for seizure and decision for additional treatment. The EEG waveforms recorded with Ceribell EEG were subsequently analyzed by three blinded epileptologists to assess the presence or absence of seizures within and outside the sonification window. Study outcomes were EEG set up time, ease of use of the device, change in clinician seizure suspicion, and change in decision to treat with AED before and after sonification.

RESULTS

Thirty-five cases of EEG sonification were performed. Mean EEG setup time was 6 ± 3 min, and time to obtain sonified EEG was significantly faster than conventional EEG (p < 0.001). One patient had non-convulsive seizure during sonification and another had rhythmic activity that was followed by seizure shortly after sonification. Change in treatment decision after sonification occurred in approximately 40% of patients and resulted in a significant net reduction in unnecessary additional treatments (p = 0.01). Ceribell EEG System was consistently rated easy to use.

CONCLUSION

The Ceribell EEG System enabled rapid acquisition of EEG in patients at risk for non-convulsive seizures and aided clinicians in their evaluation of encephalopathic ICU patients. The ease of use and speed of EEG acquisition and interpretation by EEG-untrained individuals has the potential to improve emergent clinical decision making by quickly detecting non-convulsive seizures in the ICU.

Electroencephalographic monitoring in the critically ill patient: What useful information can it contribute?

Monitoring is a crucial part of the care of the critically ill patient. It detects organ dysfunction and provides guidance on the therapeutic approach. Intensivists closely monitor the function of various organ systems, and the brain is no exception. Continuous EEG monitoring is a noninvasive and uninterrupted way of assessing cerebral cortical activity with good spatial and excellent temporal resolution. The diagnostic effectiveness of non-convulsive status epilepticus as a cause of unexplained consciousness disorder has increased the use of continuous EEG monitoring in the neurocritical care setting. However, non-convulsive status epilepticus is not the only indication for the assessment of cerebral cortical activity. This study summarizes the indications, usage and methodology of continuous EEG monitoring in the intensive care unit, with the aim of allowing practitioners to become familiarized the technique.