Interrater agreement in visual scoring of neonatal seizures based on majority voting on a web-based system: The Neoguard EEG database

Development of a stereo-EEG based seizure matching system for clinical decision making in epilepsy surgery

Objective.The proportion of patients becoming seizure-free after epilepsy surgery has stagnated. Large multi-center stereo-electroencephalography (SEEG) datasets can allow comparing new patients to past similar cases and making clinical decisions with the knowledge of how cases were treated in the past. However, the complexity of these evaluations makes the manual search for similar patients impractical. We aim to develop an automated system that electrographically and anatomically matches seizures to those in a database. Additionally, since features that define seizure similarity are unknown, we evaluate the agreement and features among experts in classifying similarity.Approach.We utilized 320 SEEG seizures from 95 consecutive patients who underwent epilepsy surgery. Eight international experts evaluated seizure-pair similarity using a four-level similarity score. As our primary outcome, we developed and validated an automated seizure matching system by employing patient data marked by independent experts. Secondary outcomes included the inter-rater agreement (IRA) and features for classifying seizure similarity.Main results.The seizure matching system achieved a median area-under-the-curve of 0.76 (interquartile range, 0.1), indicating its feasibility. Six distinct seizure similarity features were identified and proved effective: onset region, onset pattern, propagation region, duration, extent of spread, and propagation speed. Among these features, the onset region showed the strongest correlation with expert scores (Spearman's rho = 0.75,p< 0.001). Additionally, the moderate IRA confirmed the practicality of our approach with an agreement of 73.9% (7%), and Gwet's kappa of 0.45 (0.16). Further, the interoperability of the system was validated on seizures from five centers.Significance.We demonstrated the feasibility and validity of a SEEG seizure matching system across patients, effectively mirroring the expertise of epileptologists. This novel system can identify patients with seizures similar to that of a patient being evaluated, thus optimizing the treatment plan by considering the results of treating similar patients in the past, potentially improving surgery outcome.

Building an Open Source Classifier for the Neonatal EEG Background: A Systematic Feature-Based Approach From Expert Scoring to Clinical Visualization

Neonatal brain monitoring in the neonatal intensive care units (NICU) requires a continuous review of the spontaneous cortical activity, i.e., the electroencephalograph (EEG) background activity. This needs development of bedside methods for an automated assessment of the EEG background activity. In this paper, we present development of the key components of a neonatal EEG background classifier, starting from the visual background scoring to classifier design, and finally to possible bedside visualization of the classifier results. A dataset with 13,200 5-minute EEG epochs (8-16 channels) from 27 infants with birth asphyxia was used for classifier training after scoring by two independent experts. We tested three classifier designs based on 98 computational features, and their performance was assessed with respect to scoring system, pre- and post-processing of labels and outputs, choice of channels, and visualization in monitor displays. The optimal solution achieved an overall classification accuracy of 97% with a range across subjects of 81-100%. We identified a set of 23 features that make the classifier highly robust to the choice of channels and missing data due to artefact rejection. Our results showed that an automated bedside classifier of EEG background is achievable, and we publish the full classifier algorithm to allow further clinical replication and validation studies.

Levetiracetam Versus Phenobarbital for Neonatal Seizures: A Randomized Controlled Trial

BACKGROUND AND OBJECTIVES

There are no US Food and Drug Administration-approved therapies for neonatal seizures. Phenobarbital and phenytoin frequently fail to control seizures. There are concerns about the safety of seizure medications in the developing brain. Levetiracetam has proven efficacy and an excellent safety profile in older patients; therefore, there is great interest in its use in neonates. However, randomized studies have not been performed. Our objectives were to study the efficacy and safety of levetiracetam compared with phenobarbital as a first-line treatment of neonatal seizures.

METHODS

The study was a multicenter, randomized, blinded, controlled, phase IIb trial investigating the efficacy and safety of levetiracetam compared with phenobarbital as a first-line treatment for neonatal seizures of any cause. The primary outcome measure was complete seizure freedom for 24 hours, assessed by independent review of the EEGs by 2 neurophysiologists.

RESULTS

Eighty percent of patients (24 of 30) randomly assigned to phenobarbital remained seizure free for 24 hours, compared with 28% of patients (15 of 53) randomly assigned to levetiracetam (P < .001; relative risk 0.35 [95% confidence interval: 0.22-0.56]; modified intention-to-treat population). A 7.5% improvement in efficacy was achieved with a dose escalation of levetiracetam from 40 to 60 mg/kg. More adverse effects were seen in subjects randomly assigned to phenobarbital (not statistically significant).

CONCLUSIONS

In this phase IIb study, phenobarbital was more effective than levetiracetam for the treatment of neonatal seizures. Higher rates of adverse effects were seen with phenobarbital treatment. Higher-dose studies of levetiracetam are warranted, and definitive studies with long-term outcome measures are needed.

Investigating and managing neonatal seizures in the UK: an explanatory sequential mixed methods approach

Background

Neonatal seizures are difficult to diagnose and, when they are, tradition dictates first line treatment is phenobarbital. There is little data on how consultants diagnose neonatal seizures, choose when to treat or how they choose aetiological investigations or drug treatments. The purpose of this study was to assess the variation across the UK in the management of neonatal seizures and explore paediatricians’ views on their diagnosis and treatment.

Methods

An explanatory sequential mixed methods approach was used (QUAN→QUAL) with equal waiting between stages. We collected quantitative data from neonatology staff and paediatric neurologists using a questionnaire sent to neonatal units and via emails from the British Paediatric Neurology Association. We asked for copies of neonatal unit guidelines on the management of seizures. The data from questionnaires was used to identify16 consultants using semi-structured interviews. Thematic analysis was used to interpret qualitative data, which was triangulated with quantitative questionnaire data.

Results

One hundred questionnaires were returned: 47.7% thought levetiracetam was as, or equally, effective as phenobarbital; 9.2% thought it was less effective. 79.6% of clinicians had seen no side effects in neonates with levetiracetam. 97.8% of unit guidelines recommended phenobarbital first line, with wide variation in subsequent drug choice, aetiological investigations, and advice on when to start treatment. Thematic analysis revealed three themes: ‘Managing uncertainty with neonatal seizures’, ‘Moving practice forward’ and ‘Multidisciplinary team working’. Consultants noted collecting evidence on anti-convulsant drugs in neonates is problematic, and recommended a number of solutions, including collaboration to reach consensus guidelines, to reduce diagnostic and management uncertainty.

Conclusions

There is wide variation in the management of neonatal seizures and clinicians face many uncertainties. Our data has helped reveal some of the reasons for current practice and decision making. Suggestions to improve certainty include: educational initiatives to improve the ability of neonatal staff to describe suspicious events, greater use of video, closer working between neonatologists and neurologists, further research, and a national discussion to reach a consensus on a standardised approach to managing neonatal epileptic seizures.

Neonatal Seizure Detection Using Deep Convolutional Neural Networks

Identifying a core set of features is one of the most important steps in the development of an automated seizure detector. In most of the published studies describing features and seizure classifiers, the features were hand-engineered, which may not be optimal. The main goal of the present paper is using deep convolutional neural networks (CNNs) and random forest to automatically optimize feature selection and classification. The input of the proposed classifier is raw multi-channel EEG and the output is the class label: seizure/nonseizure. By training this network, the required features are optimized, while fitting a nonlinear classifier on the features. After training the network with EEG recordings of 26 neonates, five end layers performing the classification were replaced with a random forest classifier in order to improve the performance. This resulted in a false alarm rate of 0.9 per hour and seizure detection rate of 77% using a test set of EEG recordings of 22 neonates that also included dubious seizures. The newly proposed CNN classifier outperformed three data-driven feature-based approaches and performed similar to a previously developed heuristic method.

Predictive value of early EEG for seizures in neonates with hypoxic-ischemic encephalopathy undergoing therapeutic hypothermia

OBJECTIVES

To assess the prognostic significance of an early normal/mildly abnormal conventional EEG (cEEG) on seizure risk in neonates undergoing therapeutic hypothermia.

METHODS

We reviewed the video-EEG recordings from a large cohort of neonates treated with therapeutic hypothermia for hypoxic-ischemic encephalopathy from 2008 to 2017 in a single tertiary center. Continuous video-EEG was started as soon as possible (median 8.2 h) and continued throughout hypothermia and rewarming. We studied those neonates with a normal/mildly abnormal EEG during the first 24 h of monitoring.

RESULTS

A total of 331 neonates were treated with hypothermia and 323 had cEEG recordings available for review; 99 were excluded because of a moderately/severely abnormal cEEG background and/or seizure during the first 24 h of recording, and an additional eight because of early rewarming. The remaining 216 had a normal/mildly abnormal cEEG in the first 24 h. None of these patients subsequently developed seizures.

CONCLUSION

A normal/mildly abnormal cEEG during the first 24 h indicates a very low risk of subsequent seizures. This suggests that cEEG monitoring can be safely discontinued after 24 h if it has remained normal or excessively discontinuous and no seizures are detected, limiting the need for this resource-intensive and expensive tool.