Seizure Detection Algorithms in Critically Ill Children: A Comparative Evaluation

Optimizing EEG monitoring in critically ill children at risk for electroencephalographic seizures

OBJECTIVE

Strategies are needed to optimally deploy continuous EEG monitoring (CEEG) for electroencephalographic seizure (ES) identification and management due to resource limitations. We aimed to construct an efficient multi-stage prediction model guiding CEEG utilization to identify ES in critically ill children using clinical and EEG covariates.

METHODS

The largest prospective single-center cohort of 1399 consecutive children undergoing CEEG was analyzed. A four-stage model was developed and trained to predict whether a subject required additional CEEG at the conclusion of each stage given their risk of ES. Logistic regression, elastic net, random forest, and CatBoost served as candidate methods for each stage and were evaluated using cross validation. An optimal multi-stage model consisting of the top-performing stage-specific models was constructed.

RESULTS

When evaluated on a test set, the optimal multi-stage model achieved a cumulative specificity of 0.197 and cumulative F1 score of 0.326 while maintaining a high minimum cumulative sensitivity of 0.938. Overall, 11 % of test subjects with ES were removed from the model due to a predicted low risk of ES (falsely negative subjects). CEEG utilization would be reduced by 32 % and 47 % compared to performing 24 and 48 h of CEEG in all test subjects, respectively. We developed a web application called EEGLE (EEG Length Estimator) that enables straightforward implementation of the model.

CONCLUSIONS

Application of the optimal multi-stage ES prediction model could either reduce CEEG utilization for patients at lower risk of ES or promote CEEG resource reallocation to patients at higher risk for ES.

Review of Noninvasive Neuromonitoring Modalities in Children II: EEG, qEEG

Critically ill children with acute neurologic dysfunction are at risk for a variety of complications that can be detected by noninvasive bedside neuromonitoring. Continuous electroencephalography (cEEG) is the most widely available and utilized form of neuromonitoring in the pediatric intensive care unit. In this article, we review the role of cEEG and the emerging role of quantitative EEG (qEEG) in this patient population. cEEG has long been established as the gold standard for detecting seizures in critically ill children and assessing treatment response, and its role in background assessment and neuroprognostication after brain injury is also discussed. We explore the emerging utility of both cEEG and qEEG as biomarkers of degree of cerebral dysfunction after specific injuries and their ability to detect both neurologic deterioration and improvement.

Comparison of Automated Spike Detection Software in Detecting Epileptiform Abnormalities on Scalp-EEG of Genetic Generalized Epilepsy Patients

PURPOSE

Despite availability of commercial EEG software for automated epileptiform detection, validation on real-world EEG datasets is lacking. Performance evaluation of two software packages on a large EEG dataset of patients with genetic generalized epilepsy was performed.

METHODS

Three epileptologists labelled IEDs manually of EEGs from three centres. All Interictal epileptiform discharge (IED) markings predicted by two commercial software (Encevis 1.11 and Persyst 14) were reviewed individually to assess for suspicious missed markings and were integrated into the reference standard if overlooked during manual annotation during a second phase. Sensitivity, precision, specificity, and F1-score were used to assess the performance of the software packages against the adjusted reference standard.

RESULTS

One hundred and twenty-five routine scalp EEG recordings from different subjects were included (total recording time, 310.7 hours). The total epileptiform discharge reference count was 5,907 (including spikes and fragments). Encevis demonstrated a mean sensitivity for detection of IEDs of 0.46 (SD 0.32), mean precision of 0.37 (SD 0.31), and mean F1-score of 0.43 (SD 0.23). Using the default medium setting, the sensitivity of Persyst was 0.67 (SD 0.31), with a precision of 0.49 (SD 0.33) and F1-score of 0.51 (SD 0.25). Mean specificity representing non-IED window identification and classification was 0.973 (SD 0.08) for Encevis and 0.968 (SD 0.07) for Persyst.

CONCLUSIONS

Automated software shows a high degree of specificity for detection of nonepileptiform background. Sensitivity and precision for IED detection is lower, but may be acceptable for initial screening in the clinical and research setting. Clinical caution and continuous expert human oversight are recommended with all EEG recordings before a diagnostic interpretation is provided based on the output of the software.

The Spectrum of Quantitative EEG Utilization Across North America: A Cross-Sectional Survey

BACKGROUND

Continuous electroencephalography (cEEG) is commonly used for neuromonitoring in pediatric intensive care units (PICU); however, there are barriers to real-time interpretation of EEG data. Quantitative EEG (qEEG) transforms the EEG signal into time-compressed graphs, which can be displayed at the bedside. A survey was designed to understand current PICU qEEG use.

METHODS

An electronic survey was sent to the Pediatric Neurocritical Care Research Group and Pediatric Status Epilepticus Research Group, and intensivists in 16 Canadian PICUs. Questions addressed demographics, qEEG acquisition and storage, clinical use, and education.

RESULTS

Fifty respondents from 39 institutions completed the survey (response rate 53% [39 of 74 institutions]), 76% (37 of 50) from the United States and 24% (12 of 50) from Canada. Over half of the institutions (22 of 39 [56%]) utilize qEEG in their ICUs. qEEG use was associated with having a neurocritical care (NCC) service, ≥200 NCC consults/year, ≥1500 ICU admissions/year, and ≥4 ICU EEGs/day (P < 0.05 for all). Nearly all users (92% [24 of 26]) endorsed that qEEG enhanced care of children with acute neurological injury. Lack of training in qEEG was identified as a common barrier [85% (22 of 26)]. Reviewing and reporting of qEEG was not standard at most institutions. Training was required by 14% (three of 22) of institutions, and 32% (seven of 22) had established curricula.

CONCLUSIONS

ICU qEEG was used at more than half of the institutions surveyed, but review, reporting, and application of this tool remained highly variable. Although providers identify qEEG as a useful tool in patient management, further studies are needed to define clinically meaningful pediatric trends, standardize reporting, and enhance educate bedside providers.

Seizure Detection in Continuous Inpatient EEG: A Comparison of Human vs Automated Review

BACKGROUND AND OBJECTIVES

The aim of this work was to test the accuracy of Persyst commercially available automated seizure detection in critical care EEG by comparing automated seizure detections to human review in a manually reviewed cohort and on a large scale.

METHODS

Automated seizure detections (Persyst versions 12 and 13) were compared to human review in a pilot cohort of 229 seizures from 85 EEG records and then in an expanded cohort of 7,924 EEG records. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for individual seizures (pilot cohort) and for entire records (pilot and expanded cohorts). We assessed EEG features associated with the accuracy of automated seizure detections.

RESULTS

In the pilot cohort, accuracy of automated detection for individual seizures was modest (sensitivity 0.50, PPV 0.60). At the record level (did the recording contain seizures or not?), sensitivity was higher (pilot cohort 0.78, expanded cohort 0.91), PPV was low (pilot cohort 0.40, expanded cohort 0.08), and NPV was high (pilot cohort 0.88, expanded cohort 0.97). Different software versions (version 12 vs 13) performed similarly. Sensitivity was higher for records containing focal-onset seizures compared to generalized-onset seizures (0.93 vs 0.85, p = 0.012).

DISCUSSION

In critical care continuous EEG recordings, automated detection of individual seizures had rates of both false negatives and false positives that bring into question its utility as a seizure alarm in clinical practice. At the level of entire EEG records, the absence of automated detections accurately predicted EEG records without true seizures. The true value of Persyst automated seizure detection appears to lie in triaging of low-risk EEGs.

CLASSIFICATION OF EVIDENCE

This study provides Class II evidence that an automated seizure detection program cannot accurately identify EEG records that contain seizures.

Effect of Early Nutritional Assessment and Nutritional Support on Immune Function and Clinical Prognosis of Critically Ill Children

The aim of this study was to study the effect of early nutritional assessment and nutritional support on immune function and clinical prognosis of critically ill children. 90 critically ill children at the same level of severity admitted to the pediatric intensive care unit (PICU) of our hospital (June 2019-June 2020) were chosen as the research objects and were equally separated into the experimental group and the control group by the random number table method. The children in the control group were admitted to the PICU according to the routine process, and the nutritional support was provided to the malnourished ones. After admission to the PICU, the children in the experimental group were given nutritional assessment, nutritional risk screening, and nutritional support according to the screening results. The PICU stay time and total hospitalization time of the experimental group were obviously shorter than those of the control group (P < 0.05), the hospitalization expenses of the experimental group were obviously lower than those of the control group (P < 0.05), the clinical outcomes and immune function of the experimental group were obviously better than those of the control group (P < 0.05), and the nutrition indicators of the experimental group were obviously higher than those of the control group (P < 0.05). Early nutritional assessment and nutritional support can effectively improve the immune function and reduce the incidence of adverse clinical outcomes of critically ill children, which are worthy of clinical application and promotion.

Spectrograms for Seizure Detection in Critically Ill Children

SUMMARY

Electrographic seizures are common in critically ill children and a significant proportion of these seizures are nonconvulsive. There is an association between electrographic seizures and neurophysiological disturbances, worse short- and long-term neurologic outcomes, and mortality in critically ill patients. In this context, timely diagnosis and treatment of electrographic seizures in critically ill children becomes important. However, most institutions lack the resources to support round-the-clock or frequent review of continuous EEG recordings causing significant delays in seizure diagnosis. Given the current gaps in review of continuous EEG across institutions globally, use of visually simplified, time-compressed quantitative EEG trends such as spectrograms has the potential to enhance timeliness of seizure diagnosis and treatment in critically ill children.

Syndromic and non-syndromic etiologies causing neonatal hypocalcemic seizures

BACKGROUND

The diagnosis of neonatal hypocalcemic seizures (HS) in newborns is made based on clinical signs and serum calcium level. Their etiology is broad and diverse, and timely detection and initiation of treatment is essential.

METHODS

We retrospectively reviewed 1029 patients admitted to the neonatal intensive care unit. Neonatal HS were diagnosed in 16 patients, and we compared etiologies and clinical outcomes, including clinical seizures and neurodevelopment at least over 1 year old.

RESULTS

The etiologies can be broadly categorized into 5 syndromic and 11 non-syndromic neonatal HS. Syndromic neonatal HS included 3 Digeorge syndrome, 1 Kleefstra syndrome and 1 Alström syndrome. Non-syndromic neonatal HS included 8 vitamin D deficiency, 1 hypoparathyroidism, and 2 hypoxic-ischemic encephalopathy. Patients with syndromic neonatal HS were found to have worse clinical outcomes than those with nonsyndromic HS. In eight patients with vitamin D deficiency, neurodevelopment was normal. Five of five patients (100%) with syndromic HS used two or more antiseizure drugs. However, among patients with non-syndromic neonatal HS, only one of 11 (9.1%) used more than one drug (p = 0.001).

CONCLUSION

This finding highlighted that syndromic hypocalcemic seizures in newborns have worse neurodevelopmental outcomes and are more often difficult to manage, and would benefit from a genetic diagnostic approach.

Utility of Quantitative EEG for Seizure Detection in Adults

SUMMARY

Traditional review of EEG for seizure detection requires time and the expertise of a trained neurophysiologist; therefore, it is time- and resource-intensive. Quantitative EEG (qEEG) encompasses a variety of methods to make EEG review more efficient and allows for nonexpert review. Literature supports that qEEG is commonly used by neurophysiologists and nonexperts in clinical practice. In this review, the different types of qEEG trends and spectrograms used for seizure detection in adults, from basic concepts to clinical applications, are discussed. The merits and drawbacks of the most common qEEG trends are detailed. The authors detail the retrospective literature on qEEG sensitivity, specificity, and false alarm rate as interpreted by experts and nonexperts alike. Finally, the authors discuss the future of qEEG as a useful screening tool and speculate on the trajectory of future investigations in the field.

aEEG as a useful tool for neuromonitoring in critically ill children - Current evidence and knowledge gaps

AIM

Amplitude-integrated electroencephalography (aEEG) is used in children beyond neonatal age, but systematic investigations have been lacking. This mini-review summarised aEEG studies on children aged one month to 18 years, evaluated the usefulness of aEEG and identified knowledge gaps or limitations.

METHODS

We searched the PubMed database for articles published in English up to September 2020, and 23 papers were identified.

RESULTS

aEEG was frequently used to compensate for the absence of continuous full-channel EEG monitoring, particularly for detecting seizures. Interpreting background patterns was based on neonatal classifications, as reference values for older infants and children are lacking. It is possible that aEEG could predict outcomes after paediatric cardiac arrests and other conditions. Gaps in our knowledge exist with regard to normal values in healthy children and the effects of sedation on aEEG background patterns in children.

CONCLUSION

The main application of aEEG was detecting and treating paediatric seizures. Further research should determine reference values and investigate the potential to predict outcome after critical events or in acute neurological disease. It is likely that aEEG will play a role in paediatric critical care in the future.

Systematic analysis and comparison of commercial seizure-detection software

OBJECTIVE

To determine if three different commercially available seizure-detection software packages (Besa 2.0, Encevis 1.7, and Persyst 13) accurately detect seizures with high sensitivity, high specificity, and short detection delay in epilepsy patients undergoing long-term video-electroencephalography (EEG) monitoring (VEM).

METHODS

Comparison of sensitivity (detection rate), specificity (false alarm rate), and detection delay of three commercially available seizure-detection software packages in 81 randomly selected patients with epilepsy undergoing long-term VEM.

RESULTS

Detection rates on a per-patient basis were not significantly different between Besa (mean 67.6%, range 0-100%), Encevis (77.8%, 0-100%) and Persyst (81%, 0-100%; P = .059). False alarm rate (per hour) was significantly different between Besa (mean 0.7/h, range 0.01-6.2/h), Encevis (0.2/h, 0.01-0.5/h), and Persyst (0.9/h, 0.04-6.5/h; P < .001). Detection delay was significantly different between Besa (mean 30 s, range 0-431 s), Encevis (25 s, 2-163 s), and Persyst (20 s, 0-167 s; P = .007). Kappa statistics showed moderate to substantial agreement between the reference standard and each seizure-detection software (Besa: 0.47, 95% confidence interval [CI] 0.36-0.59; Encevis: 0.59, 95% CI 0.47-0.7; Persyst: 0.63, 95% CI 0.51-0.74).

SIGNIFICANCE

Three commercially available seizure-detection software packages showed similar, reasonable sensitivities on the same data set, but differed in false alarm rates and detection delay. Persyst 13 showed the highest detection rate and false alarm rate with the shortest detection delay, whereas Encevis 1.7 had a slightly lower sensitivity, the lowest false alarm rate, and longer detection delay.

Amplitude Integrated Electroencephalography and Continuous Electroencephalography Monitoring Is Crucial in High-Risk Infants and Their Findings Correlate With Neurodevelopmental Outcomes

Background: To evaluate seizure diagnosis in sick infants in the neonatal intensive care unit (NICU) based on electroencephalography (EEG) monitoring combined with amplitude integrated electroencephalography (aEEG). Methods: We retrospectively reviewed EEG and aEEG findings and determined their correlations with neurodevelopmental outcomes at the age of >1 year in 65 patients with diagnosed seizures, encephalopathy, or both. Results: Seizure identification rate was 43.1%. The rate in nonstructural groups (hypocalcemic, hypoglycemic, and genetic seizures) was 71.4%, which was higher (p < 0.05) than the rate of 35.3% of structural brain lesion group [hypoxic-ischemic encephalopathy (HIE) and congenital brain structural malformation]. The aEEG background correlating with neurodevelopmental outcomes had 70.0% positive prediction value (PPV), 65.5%% negative prediction value (NPV), 67.7% specificity, and 67.9% sensitivity (p < 0.005). The aEEG background strongly (PPV, 93.8%; p < 0.005) correlated with the outcomes in HIE. For genetic seizures, the detected rate was high. The ictal recordings for the nonstructural seizures revealed downflected on the aEEG background initially, which differed from the structural lesion. Conclusions: EEG monitoring combined with aEEG can detect seizures, facilitating early treatment. EEG changes during seizures could exhibit delta-theta waves with or without clinical seizures in patients with brain lesions. In non-structural etiologies (hypocalcemic and KCNQ2 seizures), aEEG initially exhibited lower background during seizures that could aid in differentiating these EEG changes from those of other etiologies. The aEEG background was correlated with neurodevelopmental outcome and exhibited high PPV but not NPV in neonatal HIE.