Yield of conventional and automated seizure detection methods in the epilepsy monitoring unit

Ex-SPECTing success: Predictors of successful SPECT radiotracer injection during presurgical video-EEG admissions

OBJECTIVES

To determine predictors of successful ictal single photon emission computed tomography (SPECT) injections during Epilepsy Monitoring Unit (EMU) admissions for patients undergoing presurgical evaluation for drug-resistant focal epilepsy.

METHODS

In this retrospective study, consecutive EMU admissions were analyzed at a single center between 2019 and 2021. All seizures that occurred during the admission were reviewed. "Injectable seizures" occurred during hours when the radiotracer was available. EMU-level data were analyzed to identify factors predictive of an EMU admission with a successful SPECT injection (successful admission). Seizure-level data were analyzed to identify factors predictive of an injectable seizure receiving a SPECT injection during the ictal phase (successful injection). A multivariate generalized linear model was used to identify predictive variables.

RESULTS

125 EMU admissions involving 103 patients (median 37 years, IQR 27.0-45.5) were analyzed. 38.8% of seizures that were eligible for SPECT (n = 134) were successfully injected; this represented 17.4% of all seizures (n = 298) that occurred during admission. Unsuccessful admissions were most commonly due to a lack of seizures during EMU-SPECT (19.3%) or no injectable seizures (62.3%). Successful EMU-SPECT was associated with baseline seizure frequency >1 per week (95% CI 2.1-3.0, P < 0.001) and focal PET hypometabolism (95% CI 2.0-3.7, P < 0.001). On multivariate analysis, the only factor associated with successful injection was patients being able to indicate they were having a seizure to staff (95% CI 1.0-4.4, P = 0.038).

SIGNIFICANCE

Completing a successful ictal SPECT study remains challenging. A baseline seizure frequency of >1 per week, a PET hypometabolic focus, and a patient's ability to indicate seizure onset were identified as predictors of success. These findings may assist EMUs in optimizing their SPECT protocols, patient selection, and resource allocation.

Utility of Quantitative EEG in Neurological Emergencies and ICU Clinical Practice

The electroencephalogram (EEG) is a cornerstone tool for the diagnosis, management, and prognosis of selected patient populations. EEGs offer significant advantages such as high temporal resolution, real-time cortical function assessment, and bedside usability. The quantitative EEG (qEEG) added the possibility of long recordings being processed in a compressive manner, making EEG revision more efficient for experienced users, and more friendly for new ones. Recent advancements in commercially available software, such as Persyst, have significantly expanded and facilitated the use of qEEGs, marking the beginning of a new era in its application. As a result, there has been a notable increase in the practical, real-world utilization of qEEGs in recent years. This paper aims to provide an overview of the current applications of qEEGs in daily neurological emergencies and ICU practice, and some elementary principles of qEEGs using Persyst software in clinical settings. This article illustrates basic qEEG patterns encountered in critical care and adopts the new terminology proposed for spectrogram reporting.

SzCORE: Seizure Community Open-Source Research Evaluation framework for the validation of electroencephalography-based automated seizure detection algorithms

The need for high-quality automated seizure detection algorithms based on electroencephalography (EEG) becomes ever more pressing with the increasing use of ambulatory and long-term EEG monitoring. Heterogeneity in validation methods of these algorithms influences the reported results and makes comprehensive evaluation and comparison challenging. This heterogeneity concerns in particular the choice of datasets, evaluation methodologies, and performance metrics. In this paper, we propose a unified framework designed to establish standardization in the validation of EEG-based seizure detection algorithms. Based on existing guidelines and recommendations, the framework introduces a set of recommendations and standards related to datasets, file formats, EEG data input content, seizure annotation input and output, cross-validation strategies, and performance metrics. We also propose the EEG 10-20 seizure detection benchmark, a machine-learning benchmark based on public datasets converted to a standardized format. This benchmark defines the machine-learning task as well as reporting metrics. We illustrate the use of the benchmark by evaluating a set of existing seizure detection algorithms. The SzCORE (Seizure Community Open-Source Research Evaluation) framework and benchmark are made publicly available along with an open-source software library to facilitate research use, while enabling rigorous evaluation of the clinical significance of the algorithms, fostering a collective effort to more optimally detect seizures to improve the lives of people with epilepsy.

Intervention time and adverse events in a canadian epilepsy monitoring unit: An updated audit

Background

Optimizing patient safety in the epilepsy monitoring unit (EMU) has become a topic of increasing interest. We performed an audit of our center's new single-floor EMU, assessing intervention rate (IR), intervention time (IT), and adverse events (AEs).

Methods

A prospective study was conducted on all clinical seizures of patients admitted over a one-year period at our Canadian academic tertiary care center's new single-floor EMU. This single-floor EMU was supervised by EEG technologists during daytime (similar to the old set-up) and beneficiary attendants during nighttime/weekends (versus live video feed to the central nursing station on the neurology ward previously). Among 153 admissions, 79 were analyzed, and a total of 537 seizures were reviewed to assess IR, IT, and AEs. Univariate comparisons were performed with our double-floor EMU, which we reported in a previous publication.

Results

In our new single-floor EMU, the IR was 61.1 % and overall median IT was 29.0s (19.0s-45.9s). The AE rate was 4.8 %. Compared to previously reported numbers for our old double-floor EMU (IR = 27.8 %; IT = 21.0s; AE = 1.2 %), the IR was significantly higher ((p < 0.001) but unexpectedly, the median IT was higher (p < 0.001) as well as the AE rate (p < 0.001).

Conclusion

This prospective evaluation revealed a small but non-negligible rate of complications in our EMU, higher than our prior retrospective audit. Heightened levels of supervision in our new single-floor EMU led to higher IR. This may have led to artificially longer ITs.

Patient Safety in Canadian Epilepsy Monitoring Units: A Survey of Current Practices

BACKGROUND

Guidelines on epilepsy monitoring unit (EMU) standards have been recently published. We aimed to survey Canadian EMUs to describe the landscape of safety practices and compare these to the recommendations from the new guidelines.

METHODS

A 34-item survey was created by compiling questions on EMU structure, patient monitoring, equipment, personnel, standardized protocol use, and use of injury prevention tools. The questionnaire was distributed online to 24 Canadian hospital centers performing video-EEG monitoring (VEM) in EMUs. Responses were tabulated and descriptively summarized.

RESULTS

In total, 26 EMUs responded (100% response rate), 50% of which were adult EMUs. EMUs were on average active for 23.4 years and had on average 3.6 beds. About 81% of respondents reported having a dedicated area for VEM, and 65% reported having designated EMU beds. Although a video monitoring station was available in 96% of EMUs, only 48% of EMUs provided continuous observation of patients (video and/or physical). A total of 65% of EMUs employed continuous heart monitoring. The technologist-to-patient ratio was 1:1-2 in 52% of EMUs during the day. No technologist supervision was most often reported in the evening and at night. Nurse-to-EMU-patient ratio was mostly 1:1-4 independent of the time of day. Consent forms were required before admission in 27% of EMUs.

CONCLUSION

Canadian EMUs performed decently in terms of there being dedicated space for VEM, continuous heart monitoring, and adequate nurse-to-patient ratios. Other practices were quite variable, and adjustments should be made on a case-by-case basis to adhere to the latest guidelines.

Epilepsy monitoring unit practices and safety among NAEC epilepsy centers: A census survey

OBJECTIVE

An epilepsy monitoring unit (EMU) is a specialized unit designed for capturing and characterizing seizures and other paroxysmal events with continuous video electroencephalography (vEEG). Nearly 260 epilepsy centers in the United States are accredited by the National Association of Epilepsy Centers (NAEC) based on adherence to specific clinical standards to improve epilepsy care, safety, and quality. This study examines EMU staffing, safety practices, and reported outcomes.

METHOD

We analyzed NAEC annual report data and results from a supplemental survey specific to EMU practices reported in 2019 from 341 pediatric or adult center directors. Data on staffing, resources, safety practices and complications were collated with epilepsy center characteristics. We summarized using frequency (percentage) for categorical variables and median (inter-quartile range) for continuous variables. We used chi-square or Fisher's exact tests to compare staff responsibilities.

RESULTS

The supplemental survey response rate was 100%. Spell classification (39%) and phase 1 testing (28%) were the most common goals of the 91,069 reported admissions. The goal ratio of EEG technologist to beds of 1:4 was the most common during the day (68%) and off-hours (43%). Compared to residents and fellows, advanced practice providers served more roles in the EMU at level 3 or pediatric-only centers. Status epilepticus (SE) was the most common reported complication (1.6% of admissions), while cardiac arrest occurred in 0.1% of admissions.

SIGNIFICANCE

EMU staffing and safety practices vary across US epilepsy centers. Reported complications in EMUs are rare but could be further reduced, such as with more effective treatment or prevention of SE. These findings have potential implications for improving EMU safety and quality care.

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.

Automated spike and seizure detection: Are we ready for implementation?

OBJECTIVE

Automated detection of spikes and seizures has been a subject of research for several decades now. There have been important advances, yet automated detection in EMU (Epilepsy Monitoring Unit) settings has not been accepted as standard practice. We intend to implement this software at our EMU and so carried out a qualitative study to identify factors that hinder ('barriers') and facilitate ('enablers') implementation.

METHOD

Twenty-two semi-structured interviews were conducted with 14 technicians and neurologists involved in recording and reporting EEGs and eight neurologists who receive EEG reports in the outpatient department. The study was reported according to the Consolidated Criteria for Reporting Qualitative Studies (COREQ).

RESULTS

We identified 14 barriers and 14 enablers for future implementation. Most barriers were reported by technicians. The most prominent barrier was lack of trust in the software, especially regarding seizure detection and false positive results. Additionally, technicians feared losing their EEG review skills or their jobs. Most commonly reported enablers included potential efficiency in the EEG workflow, the opportunity for quantification of EEG findings and the willingness to try the software.

CONCLUSIONS

This study provides insight into the perspectives of users and offers recommendations for implementing automated spike and seizure detection in EMUs.

Moving the field forward: detection of epileptiform abnormalities on scalp electroencephalography using deep learning—clinical application perspectives

The application of deep learning approaches for the detection of interictal epileptiform discharges is a nascent field, with most studies published in the past 5 years. Although many recent models have been published demonstrating promising results, deficiencies in descriptions of data sets, unstandardized methods, variation in performance evaluation and lack of demonstrable generalizability have made it difficult for these algorithms to be compared and progress to clinical validity. A few recent publications have provided a detailed breakdown of data sets and relevant performance metrics to exemplify the potential of deep learning in epileptiform discharge detection. This review provides an overview of the field and equips computer and data scientists with a synopsis of EEG data sets, background and epileptiform variation, model evaluation parameters and an awareness of the performance metrics of high impact and interest to the trained clinical and neuroscientist EEG end user. The gold standard and inter-rater disagreements in defining epileptiform abnormalities remain a challenge in the field, and a hierarchical proposal for epileptiform discharge labelling options is recommended. Standardized descriptions of data sets and reporting metrics are a priority. Source code-sharing and accessibility to public EEG data sets will increase the rigour, quality and progress in the field and allow validation and real-world clinical translation.

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.

Dynamic training of a novelty classifier algorithm for real-time detection of early seizure onset

OBJECTIVE

To develop an adaptive framework for seizure detection in real-time that is practical to use in the Epilepsy Monitoring Unit (EMU) as a warning signal, and whose output helps characterize epileptiform activity.

METHODS

Our algorithm was tested on intracranial EEG from epilepsy patients admitted to the EMU for presurgical evaluation. Our framework uses a one-class Support Vector Machine (SVM) that is being trained dynamically according to past activity in all available channels to classify the novelty of the current activity. In this study we compared multiple configurations using a one-class SVM to assess if there is significance over specific neural features or electrode locations.

RESULTS

Our results show that the algorithm reaches a sensitivity of 87% for early-onset seizure detection and of 97.7% as a generic seizure detection.

CONCLUSIONS

Our algorithm is capable of running in real-time and achieving a high performance for early seizure-onset detection with a low false positive rate and robustness in detection of different type of seizure-onset patterns.

SIGNIFICANCE

This algorithm offers a solution to warning systems in the EMU as well as a tool for seizure characterization during post-hoc analysis of intracranial EEG data for surgical resection of the epileptogenic network.

Standardization of seizure response times and data collection in an epilepsy monitoring unit

OBJECTIVE

We sought to improve seizure response times in the epilepsy monitoring unit, improve the accuracy and reliability of seizure response time data collection, and develop a standardized and automated approach for seizure response data collection in the EMU.

METHODS

We used Quality Improvement (QI) methodology to understand the EMU workflow involved in responding to seizures (a process map); to create a theory of change that stated the desired aim, potential drivers/barriers and interventions (i.e., key driver diagram) and perform iterative interventions to address some of the drivers plan-do-study-act (PDSA) cycles. We performed three PDSA cycles with a focus on improving the seizure alert system in our EMU. Adjustments were made to the methodology as it became clear that this was a systems issue, and our project would need to focus on improving the system rather than iteratively improving a functioning (stable) system.

RESULTS

Over a 6-month period, 252 seizure response times were recorded and analyzed. We performed 3 interventions. The first was initiating twice monthly meetings with nursing and EEG techs to discuss the project and provide feedback on response times. The second was the implementation of a new Hill-Rom seizure alert system to reduce alert times and automate data tracking. The third was implementing a new alert deactivation system to reduce variability in the data. Following these 3 interventions, variation, and data collection methods were improved while also maintaining improvements in seizure response times.

SIGNIFICANCE

We identified and implemented an alert system in our EMU which led to more efficient and accurate data collection while maintaining improved response times that resulted from the first intervention. This lays the groundwork for future QI initiatives and has created a framework for standardizing seizure response time recording and data collection that can be replicated at other centers with similar infrastructure, personnel and workflows.

Automated seizure detection in an EMU setting: Are software packages ready for implementation?

PURPOSE

We assessed whether automated detection software, combined with live observation, enabled reliable seizure detection using three commercial software packages: Persyst, Encevis and BESA.

METHODS

Two hundred and eighty-six prolonged EEG records of individuals aged 16-86 years, collected between August 2019 and January 2020, were retrospectively processed using all three packages. The reference standard included all seizures mentioned in the clinical report supplemented with true detections made by the software and not previously detected by clinical physiologists. Sensitivity was measured for offline review by clinical physiologists and software seizure detection, both in combination with live monitoring in an EMU setting, for all three software packages at record and seizure level.

RESULTS

The database contained 249 seizures in 64 records. The sensitivity of seizure detection was 98% for Encevis and Persyst, and 95% for BESA, when a positive results was defined as detection at least one of the seizures occurring within an individual record. When positivity was defined as recognition of all seizures, sensitivity was 93% for Persyst, 88% for Encevis and 84% for BESA. Clinical physiologists' review had a sensitivity of 100% at record level and 98% at seizure level. The median false positive rate per record was 1.7 for Persyst, 2.4 for BESA and 5.5 for Encevis per 24 h.

CONCLUSION

Automated seizure detection software does not perform as well as technicians do. However, it can be used in an EMU setting when the user is aware of its weaknesses. This assessment gives future users helpful insight into these strengths and weaknesses. The Persyst software performs best.

Seizure Detection: Interreader Agreement and Detection Algorithm Assessments Using a Large Dataset

PURPOSE

To compare the seizure detection performance of three expert humans and two computer algorithms in a large set of epilepsy monitoring unit EEG recordings.

METHODS

One hundred twenty prolonged EEGs, 100 containing clinically reported EEG-evident seizures, were evaluated. Seizures were marked by the experts and algorithms. Pairwise sensitivity and false-positive rates were calculated for each human-human and algorithm-human pair. Differences in human pairwise performance were calculated and compared with the range of algorithm versus human performance differences as a type of statistical modified Turing test.

RESULTS

A total of 411 individual seizure events were marked by the experts in 2,805 hours of EEG. Mean, pairwise human sensitivities and false-positive rates were 84.9%, 73.7%, and 72.5%, and 1.0, 0.4, and 1.0/day, respectively. Only the Persyst 14 algorithm was comparable with humans-78.2% and 1.0/day. Evaluation of pairwise differences in sensitivity and false-positive rate demonstrated that Persyst 14 met statistical noninferiority criteria compared with the expert humans.

CONCLUSIONS

Evaluating typical prolonged EEG recordings, human experts had a modest level of agreement in seizure marking and low false-positive rates. The Persyst 14 algorithm was statistically noninferior to the humans. For the first time, a seizure detection algorithm and human experts performed similarly.

Integrating old and new complexity measures toward automated seizure detection from long-term video EEG recordings

Automated seizure detection in long-term video-EEG recordings is far from being integrated into common clinical practice. Here, we leverage classical and state-of-the-art complexity measures to robustly and automatically detect seizures from scalp recordings. Brain activity is scored through eight features, encompassing traditional time domain and novel measures of recurrence. A binary classification algorithm tailored to treat unbalanced dataset is used to determine whether a time window is ictal or non-ictal from its features. The application of the algorithm on a cohort of ten adult patients with focal refractory epilepsy indicates sensitivity, specificity, and accuracy of 90%, along with a true alarm rate of 95% and less than four false alarms per day. The proposed approach emphasizes ictal patterns against noisy background without the need of data preprocessing. Finally, we benchmark our approach against previous studies on two publicly available datasets, demonstrating the good performance of our algorithm.

•

Complexity measures are formulated to enhance classical time-domain statistics of EEG

•

The detection algorithm does not need ad-hoc data preprocessing to address artifacts

•

Focal seizures are detected 95% of the time with less than four false alarms per day

•

The approach offers a visual representation of a seizure as a time-evolving network

Intervention Time and Adverse Events in a Canadian Epilepsy Monitoring Unit

BACKGROUND

Intervention time (IT) in response to seizures and adverse events (AEs) have emerged as key elements in epilepsy monitoring unit (EMU) management. We performed an audit of our EMU, focusing on IT and AEs.

METHODS

We performed a retrospective study on all clinical seizures of admissions over a 1-year period at our Canadian academic tertiary care center's EMU. This EMU was divided in two subunits: a daytime three-bed epilepsy department subunit (EDU) supervised by EEG technicians and a three-bed neurology ward subunit (NWU) equipped with video-EEG where patients were transferred to for nights and weekends, under nursing supervision. Among 124 admissions, 58 were analyzed. A total of 1293 seizures were reviewed to determine intervention occurrence, IT, and AE occurrence. Seizures occurring when the staff was present at bedside at seizure onset were analyzed separately.

RESULTS

Median IT was 21.0 (11.0-40.8) s. The EDU, bilateral tonic-clonic seizures (BTCS), and the presence of a warning signal were associated with increased odds of an intervention taking place. The NWU, BTCS, and seizure rank (seizures were chronologically ordered by the patient for each subunit) were associated with longer ITs. Bedside staff presence rate was higher in the EDU than in the NWU (p < 0.001). AEs occurred in 19% of admissions, with no difference between subunits. AEs were more frequent in BTCS than in other seizure types (p = 0.001).

CONCLUSION

This study suggests that close monitoring by trained staff members dedicated to EMU patients is key to optimize safety. AE rate was high, warranting corrective measures.

Using sampled visual EEG review in combination with automated detection software at the EMU

PURPOSE

Complete visual review of prolonged video-EEG recordings at an EMU (Epilepsy Monitoring Unit) is time consuming and can cause problems in times of paucity of educated personnel. In this study we aimed to show non inferiority for electroclinical diagnosis using sampled review in combination with EEG analysis softreferware (P13 software, Persyst Corporation), in comparison to complete visual review.

METHOD

Fifty prolonged video-EEG recordings in adults were prospectively evaluated using sampled visual EEG review in combination with automated detection software of the complete EEG record. Visually assessed samples consisted of one hour during wakefulness, one hour during sleep, half an hour of wakefulness after wake-up and all clinical events marked by the individual and/or nurses. The final electro-clinical diagnosis of this new review approach was compared with the electro-clinical diagnosis after complete visual review as presently used.

RESULTS

The electro-clinical diagnosis based on sampled visual review combined with automated detection software did not differ from the diagnosis based on complete visual review. Furthermore, the detection software was able to detect all records containing epileptiform abnormalities and epileptic seizures.

CONCLUSION

Sampled visual review in combination with automated detection using Persyst 13 is non-inferior to complete visual review for electroclinical diagnosis of prolonged video-EEG at an EMU setting, which makes this approach promising.