Strategies for non-EEG seizure detection and timing for alerting and interventions with tonic-clonic seizures

Learning Robust Representations of Tonic-Clonic Seizures With Cyclic Transformer

Tonic-clonic seizures (TCSs) pose a significant risk for sudden unexpected death in epilepsy (SUDEP). Previous research has highlighted the potential of multimodal wearable seizure detection systems in accurately detecting TCSs through continuous monitoring, enabling timely alarms and potentially preventing SUDEP. However, such multimodal systems carry a higher risk of sensor malfunction. In this paper, we propose a cyclic transformer approach to address these challenges. The cyclic transformer learns a robust representation by performing circular modal translations between the source and target modalities. It leverages back-translation as regularization technique to enhance the discriminative power of the learned representation. Notably, the proposed cyclic transformer is trained on paired multimodal data but requires only a single source modality during deployment. This characteristic ensures the robustness of the cyclic transformer to perturbations or missing information in the target modality. Experimental results demonstrate that the proposed cyclic transformer achieves competitive performance compared with existing multimodal systems. While both approaches were trained using EEG and EMG data, the cyclic transformer exclusively employs EEG data for testing, diverging from the state-of-the-art's utilization of both EEG and EMG data during test. This showcases the effectiveness of the cyclic transformer in multimodal TCSs detection, offering a promising approach for enhancing the accuracy and robustness of seizure detection systems while mitigating the risks associated with sensor malfunction.

A systematic review of the literature reporting on remote monitoring epileptic seizure detection devices

BACKGROUND

Early detection and alert notification of an impending seizure for people with epilepsy have the potential to reduce Sudden Unexpected Death in Epilepsy (SUDEP). Current remote monitoring seizure detection devices for people with epilepsy are designed to support real-time monitoring of their vital health parameters linked to seizure alert notification. An understanding of the rapidly growing literature on remote seizure detection devices is essential to address the needs of people with epilepsy and their carers.

AIM

This review aims to examine the technical characteristics, device performance, user preference, and effectiveness of remote monitoring seizure detection devices.

METHODOLOGY

A systematic review referenced to PRISMA guidelines was used.

RESULTS

A total of 1095 papers were identified from the initial search with 30 papers included in the review. Sixteen non-invasive remote monitoring seizure detection devices are currently available. Such seizure detection devices were found to have inbuilt intelligent sensor functionality to monitor electroencephalography, muscle movement, and accelerometer-based motion movement for detecting seizures remotely. Current challenges of these devices for people with epilepsy include skin irritation due to the type of patch electrode used and false alarm notifications, particularly during physical activity. The tight-fitted accelerometer-type devices are reported as uncomfortable from a wearability perspective for long-term monitoring. Also, continuous recording of physiological signals and triggering alert notifications significantly reduce the battery life of the devices. The literature highlights that 3.2 out of 5 people with epilepsy are not using seizure detection devices because of the cost and appearance of the device.

CONCLUSION

Seizure detection devices can potentially reduce morbidity and mortality for people with epilepsy. Therefore, further collaboration of clinicians, technical experts, and researchers is needed for the future development of these devices. Finally, it is important to always take into consideration the expectations and requirements of people with epilepsy and their carers to facilitate the next generation of remote monitoring seizure detection devices.

Getting Under the Skin of Seizure Monitoring: A Subcutaneous EEG Tool to Keep a Tally Over the Long Haul

Detecting Temporal Lobe Seizures in Ultra Long-Term Subcutaneous EEG Using Algorithm-Based Data Reduction Remvig LS, Duun-Henriksen J, Fürbass F, Hartmann M, Viana PF, Kappel Overby AM, Weisdorf S, Richardson MP, Beniczky S, Kjaer TW. Clin Neurophysiol . 2022;142:86-93. doi:10.1016/j.clinph.2022.07.504 . PMID: 35987094 Objective: Ultra long-term monitoring with subcutaneous EEG (sqEEG) offers objective outpatient recording of electrographic seizures as an alternative to self-reported epileptic seizure diaries. This methodology requires an algorithm-based automatic seizure detection to indicate periods of potential seizure activity to reduce the time spent on visual review. The objective of this study was to evaluate the performance of a sqEEG-based automatic seizure detection algorithm. Methods: A multicenter cohort of subjects using sqEEG were analyzed, including nine people with epilepsy (PWE) and 12 healthy subjects, recording a total of 965 days. The automatic seizure detections of a deep-neural-network algorithm were compared to annotations from three human experts. Results: Data reduction ratios were 99.6% in PWE and 99.9% in the control group. The cross-PWE sensitivity was 86% (median 80%, range 69-100% when PWE were evaluated individually), and the corresponding median false detection rate was 2.4 detections per 24 hours (range: 2.0-13.0). Conclusions: Our findings demonstrated that step one in a sqEEG-based semi-automatic seizure detection/review process can be performed with high sensitivity and clinically applicable specificity. Significance: Ultra long-term sqEEG bears the potential of improving objective seizure quantification.

Wrist-worn smartwatch and predictive models for seizures

OBJECTIVE

This study was undertaken to describe extracerebral biosignal characteristics of overall and various seizure types as compared with baseline physical activities using multimodal devices (Empatica E4); develop predictive models for overall and each seizure type; and assess diagnostic performance of each model.

METHODS

We prospectively recruited patients with focal epilepsy who were admitted to the epilepsy monitoring unit for presurgical evaluation during January to December 2020. All study participants were simultaneously applied gold standard long-term video-electroencephalographic (EEG) monitoring and an index test, E4. Two certified epileptologists independently determined whether captured events were seizures and then indicated ictal semiology and EEG information. Both were blind to multimodal biosignal findings detected by E4. Biosignals during 5-min epochs of both seizure events and baseline were collected and compared. Predictive models for occurrence overall and of each seizure type were developed using a generalized estimating equation. Diagnostic performance of each model was then assessed.

RESULTS

Thirty patients had events recorded and were recruited for analysis. One hundred eight seizure events and 120 baseline epochs were collected. Heart rate (HR), acceleration (ACC), and electrodermal activity (EDA) but not temperature were significantly elevated during seizures. Cluster analysis showed trends of greatest elevation of HR and ACC in bilateral tonic-clonic seizures (BTCs), as compared with non-BTCs and isolated auras. HR and ACC were independent predictors for overall seizure types, BTCs, and non-BTCs, whereas only HR was a predictor for isolated aura. Diagnostic performance including sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve of the predictive model for overall seizures were 77.78%, 60%, and .696 (95% confidence interval = .628-.764), respectively.

SIGNIFICANCE

Multimodal extracerebral biosignals (HR, ACC, EDA) detected by a wrist-worn smartwatch can help differentiate between epileptic seizures and normal physical activities. It would be worthwhile to implement our predictive algorithms in commercial seizure detection devices. However, larger studies to externally validate our predictive models are required.

Assessment of an under-mattress sensor as a seizure detection tool in an adult epilepsy monitoring unit

OBJECTIVE

Because of SUDEP (Sudden and unexpected death in epilepsy) and other direct consequences of generalized tonic-clonic seizures, the use of efficient seizure detection tool may be helpful for patients, relatives and caregivers. We aimed to evaluate an under-mattress detection tool (EMFIT®) in real-life hospital conditions, in particular its sensitivity and false alarm rate (FAR), as well as its impact on patient care.

METHODS

We carried out a retrospective study on a cohort of patients with epilepsy admitted between September 2017 and June 2021 to Amiens University Hospital for a video-EEG of at least 24 h, during which at least one epileptic seizure was recorded. All video-EEGs records were analyzed visually in order to assess the sensitivity of the under-mattress tool (triggering of the alarm) and to classify the seizure type (convulsive/non convulsive). We also considered whether nurses intervened during the seizure, and the time of their intervention if applicable. An additional prospective survey was conducted over 272 days to analyze the FAR of the tool.

RESULTS

A total of 220 seizures were included in the study, from 55 patients, including 23 convulsive seizures from 15 patients and 197 non-convulsive seizures. Sensitivity for convulsive seizure detection was 69.6%. As expected, none of the non-convulsive seizures was detected. The false alarm rate was 0.007/day. Median trigger time was 74 s, decreasing to 5 s for generalized tonic-clonic seizure. The frequency of nurses' intervention during convulsive seizures was significantly greater in case of the alarm triggering (100% vs 57%, p<0.02).

SIGNIFICANCE

These results suggest that EMFIT® sensor is able to detect convulsive seizures with good sensitivity and low FAR, and allows caregivers to intervene more often in the event of a nocturnal seizure. This would be an interesting complementary tool to better secure the patients with epilepsy during hospitalization or at home.

Toward automated prediction of sudden unexpected death in epilepsy

Sudden unexpected death in epilepsy (SUDEP) is a devastating yet overlooked complication of epilepsy. The rare and complex nature of SUDEP makes it challenging to study. No prediction or prevention of SUDEP is currently available in a clinical setting. In the past decade, significant advances have been made in our knowledge of the pathophysiologic cascades that lead to SUDEP. In particular, studies of brain, heart, and respiratory functions in both human patients at the epilepsy monitoring unit and animal models during fatal seizures provide critical information to integrate computational tools for SUDEP prediction. The rapid advances in automated seizure detection and prediction algorithms provide a fundamental framework for their adaption in predicting SUDEP. If a SUDEP can be predicted, then there will be a potential for medical intervention to be administered, either by their caregivers or via an implanted device automatically delivering electrical stimulation or medication, and finally save lives from fatal seizures. This article presents recent developments of SUDEP studies focusing on the pathophysiologic basis of SUDEP and computational implications of machine learning techniques that can be adapted and extended for SUDEP prediction. This article also discusses some novel ideas for SUDEP prediction and rescue including principal component analysis and closed-loop intervention.

Sudden Unexpected Death in Epilepsy: Pathogenesis, Risk Factors, and Prevention

Sudden unexpected death in epilepsy (SUDEP) is a tragic and unexpected cause of death in patients with a known diagnosis of epilepsy. It occurs in up to 6.3 to 9.3/1,000 patients with drug-resistant epilepsy. The main three risk factors associated with SUDEP are the presence of generalized tonic-clonic seizures, the presence of a seizure in the past year, and an intellectual disability. There are several mechanisms that can result in SUDEP. The most likely sequence of events appears to be a convulsive seizure, overactivation of the autonomic nervous system, cardiorespiratory dysfunction, and death. While the risk of SUDEP is relatively high in patients with drug-resistant epilepsy, studies indicate that more than 50% of patients and caregivers are unaware of the diagnosis. Counseling about the diagnosis and preventative measures at the time of diagnosis is important. There are numerous interventions that may reduce the risk of SUDEP, including conservative measures such as nocturnal surveillance with a bed partner (where applicable) and automated devices. Optimizing seizure control with antiseizure medications and surgical interventions can result in a reduced risk of SUDEP.

Autonomic manifestations of epilepsy: emerging pathways to sudden death?

Epileptic networks are intimately connected with the autonomic nervous system, as exemplified by a plethora of ictal (during a seizure) autonomic manifestations, including epigastric sensations, palpitations, goosebumps and syncope (fainting). Ictal autonomic changes might serve as diagnostic clues, provide targets for seizure detection and help us to understand the mechanisms that underlie sudden unexpected death in epilepsy (SUDEP). Autonomic alterations are generally more prominent in focal seizures originating from the temporal lobe, demonstrating the importance of limbic structures to the autonomic nervous system, and are particularly pronounced in focal-to-bilateral and generalized tonic-clonic seizures. The presence, type and severity of autonomic features are determined by the seizure onset zone, propagation pathways, lateralization and timing of the seizures, and the presence of interictal autonomic dysfunction. Evidence is mounting that not all autonomic manifestations are linked to SUDEP. In addition, experimental and clinical data emphasize the heterogeneity of SUDEP and its infrequent overlap with sudden cardiac death. Here, we review the spectrum and diagnostic value of the mostly benign and self-limiting autonomic manifestations of epilepsy. In particular, we focus on presentations that are likely to contribute to SUDEP and discuss how wearable devices might help to prevent SUDEP.

Seizure detection using wearable sensors and machine learning: Setting a benchmark

OBJECTIVE

Tracking seizures is crucial for epilepsy monitoring and treatment evaluation. Current epilepsy care relies on caretaker seizure diaries, but clinical seizure monitoring may miss seizures. Wearable devices may be better tolerated and more suitable for long-term ambulatory monitoring. This study evaluates the seizure detection performance of custom-developed machine learning (ML) algorithms across a broad spectrum of epileptic seizures utilizing wrist- and ankle-worn multisignal biosensors.

METHODS

We enrolled patients admitted to the epilepsy monitoring unit and asked them to wear a wearable sensor on either their wrists or ankles. The sensor recorded body temperature, electrodermal activity, accelerometry (ACC), and photoplethysmography, which provides blood volume pulse (BVP). We used electroencephalographic seizure onset and offset as determined by a board-certified epileptologist as a standard comparison. We trained and validated ML for two different algorithms: Algorithm 1, ML methods for developing seizure type-specific detection models for nine individual seizure types; and Algorithm 2, ML methods for building general seizure type-agnostic detection, lumping together all seizure types.

RESULTS

We included 94 patients (57.4% female, median age = 9.9 years) and 548 epileptic seizures (11 066 h of sensor data) for a total of 930 seizures and nine seizure types. Algorithm 1 detected eight of nine seizure types better than chance (area under the receiver operating characteristic curve [AUC-ROC] = .648-.976). Algorithm 2 detected all nine seizure types better than chance (AUC-ROC = .642-.995); a fusion of ACC and BVP modalities achieved the best AUC-ROC (.752) when combining all seizure types together.

SIGNIFICANCE

Automatic seizure detection using ML from multimodal wearable sensor data is feasible across a broad spectrum of epileptic seizures. Preliminary results show better than chance seizure detection. The next steps include validation of our results in larger datasets, evaluation of the detection utility tool for additional clinical seizure types, and integration of additional clinical information.

Electrographic seizure monitoring with a novel, wireless, single-channel EEG sensor

Objective

Recording seizures using personal seizure diaries can be challenging during everyday life and many seizures are missed or mis-reported. People living with epilepsy could benefit by having a more accurate and objective wearable EEG system for counting seizures that can be used outside of the hospital. The objective of this study was to (1) determine which seizure types can be electrographically recorded from the scalp below the hairline, (2) determine epileptologists' ability to identify electrographic seizures from single-channels extracted from full-montage wired-EEG, and (3) determine epileptologists' ability to identify electrographic seizures from Epilog, a wireless single-channel EEG sensor.

Methods

Epilog sensors were worn concurrently during epilepsy monitoring unit (EMU) monitoring. During standard-of-care review, epileptologists were asked if the electrographic portion of the seizure was visible on single channels of wired electrodes at locations proximal to Epilog sensors, and if focal-onset, which electrode was closest to the focus. From these locations, single channels of EEG extracted from wired full-montage EEG and the proximal Epilog sensor were presented to 3 blinded epileptologists along with markers for when known seizures occurred (taken from the standard-of-care review). Control segments at inter-ictal times were included as control. The epileptologists were asked whether a seizure event was visible in the single channel EEG record at or near the marker.

Results

A total of 75 seizures were recorded from 22 of 40 adults that wore Epilog during their visit to the EMU. Epileptologists were able to visualize known seizure activity on at least one of the wired electrodes proximal to Epilog sensors for all seizure events. Epileptologists accurately identified seizures in 71% of Epilog recordings and 84% of single-channel wired recordings and were 92% accurate identifying seizures with Epilog when those seizures ended in a clinical convulsion compared to those that did not (>55%).

Conclusions

Epileptologists are able to visualize seizure activity on single-channels of EEG at locations where Epilog sensors are easily placed on the scalp below hairline. Manual review of seizure annotations can be done quickly and accurately (>70% TP and >98% PPV) on single-channel EEG data. Reviewing single-channel EEG is more accurate than what has been reported in the literature on self-reporting seizures in seizure diaries, the current standard of care for seizure counting outside of the EMU.

Significance

Wearable EEG will be important for seizure monitoring outside of the hospital. Epileptologists can accurately identify seizures in single-channel EEG, better than patient self-reporting in diaries based on the literature. Automated or semi-automated seizure detection on single channels of EEG could be used in the future to objectively count seizures to complement the standard of care outside of the EMU without the overt burden upon epileptologist review.

Forecasting Seizure Storms Using Epilepsy Wristband Sensors

Machine Learning From Wristband Sensor Data for Wearable, Noninvasive Seizure Forecasting

Meisel C, El Atrache R, Jackson M, Schubach S, Ufongene C, Loddenkemper T. Epilepsia. 2020;61(12):2653-2666. doi:10.1111/epi.16719

Objective:

Seizure forecasting may provide patients with timely warnings to adapt their daily activities and help clinicians deliver more objective, personalized treatments. Although recent work has convincingly demonstrated that seizure risk assessment is in principle possible, these early approaches relied largely on complex, often invasive setups including intracranial electrocorticography, implanted devices, and multichannel electroencephalography and required patient-specific adaptation or learning to perform optimally, all of which limit translation to broad clinical application. To facilitate broader adaptation of seizure forecasting in clinical practice, noninvasive, easily applicable techniques that reliably assess seizure risk without much prior tuning are crucial. Wristbands that continuously record physiological parameters, including electrodermal activity, body temperature, blood volume pulse, and actigraphy, may afford monitoring of autonomous nervous system function and movement relevant for such a task, hence minimizing potential complications associated with invasive monitoring and avoiding stigma associated with bulky external monitoring devices on the head.

Methods:

Here, we applied deep learning on multimodal wristband sensor data from 69 patients with epilepsy (total duration >2311 hours, 452 seizures) to assess its capability to forecast seizures in a statistically significant way.

Results:

Using a leave-one-subject-out cross-validation approach, we identified better-than-chance predictability in 43% of the patients. Time-matched seizure surrogate data analyses indicated forecasting not to be driven simply by time of day or vigilance state. Prediction performance peaked when all sensor modalities were used and did not differ between generalized and focal seizure types but generally increased with the size of the training data set, indicating potential further improvement with larger data sets in the future.

Significance:

Collectively, these results show that statistically significant seizure risk assessments are feasible from easy-to-use, noninvasive wearable devices without the need of patient-specific training or parameter optimization.

Evaluation of an Activity Tracker to Detect Seizures Using Machine Learning

Currently, the tracking of seizures is highly subjective, dependent on qualitative information provided by the patient and family instead of quantifiable seizure data. Usage of a seizure detection device to potentially detect seizure events in a population of epilepsy patients has been previously done. Therefore, we chose the Fitbit Charge 2 smart watch to determine if it could detect seizure events in patients when compared to continuous electroencephalographic (EEG) monitoring for those admitted to an epilepsy monitoring unit. A total of 40 patients were enrolled in the study that met the criteria between 2015 and 2016. All seizure types were recorded. Twelve patients had a total of 53 epileptic seizures. The patient-aggregated receiver operating characteristic curve had an area under the curve of 0.58 [0.56, 0.60], indicating that the neural network models were generally able to detect seizure events at an above-chance level. However, the overall low specificity implied a false alarm rate that would likely make the model unsuitable in practice. Overall, the use of the Fitbit Charge 2 activity tracker does not appear well suited in its current form to detect epileptic seizures in patients with seizure activity when compared to data recorded from the continuous EEG.

A clinical evaluation of a novel algorithm in the reliable detection of epileptic seizures

PURPOSE

Undetected and prolonged epileptic seizures can result in hypoxic brain damage or death and occur most often when the victim is in bed alone or unsupervised. Sudden unexpected death in epilepsy may not always be preventable but it is believed that timely assistance with rescue medication and body re-positioning may overcome respiratory compromise in some cases. A novel algorithm based on a real time moving 9 s epoch, calculating 25 % percentage heart rate change and/or an oxygen saturation trigger level of <85 % was developed using photoplethysmography and incorporated into a prototype data storage device.

METHODS

The algorithm was clinically evaluated in this multicentre trial in the detection of clinically significant epileptic seizures. A range of epileptic seizures and normal physiological events were recorded and classified by reference standard EEG Videotelemetry and time-synchronised event data recorded by the prototype device incorporating the pre-specified cut-off points prospectively and retrospective analysis of all events.

RESULTS

119 participants who were attending electroencephalographic (EEG) videotelemetry as part of their clinical management of their epilepsy consented to take part in the trial. 683 epileptic seizures (77 clinically significant seizures) and 2648 normal physiological events were captured. When using pre-specified cut-off point 25 % heart rate change and/or oxygen desaturation <85 % on the basis of one/other, the device showed a sensitivity of 87 % for detecting clinically significant seizures. False Alarm Rate 4.5 (24 h FAR), detection latency of 58 s using heart rate percentage change.

CONCLUSIONS

The results indicate that the novel algorithm can be used in detecting clinically significant seizures.

Postictal stertor: Associations with focal and bilateral seizure types

OBJECTIVE

The objective of the present study was to determine the association between respiratory stertor and focal and bilateral seizure types.

METHODS

We characterized ictal and postictal behaviors during symmetric bilateral tonic-clonic (TC) and asymmetric TC seizures in the Johns Hopkins University (JHU) epilepsy monitoring unit, comparing these to focal unaware seizures. We measured the presence and duration of postictal stertorous respirations, postictal generalized electroencephalographic suppression (PGES), immobility/motor dysfunction, and encephalopathy and determined their associations and relationship to seizure types.

RESULTS

In initial seizures recorded in 80 consecutive patients, bilateral symmetric TC seizures (N = 35) were strongly associated with PGES (97%, p < 0.001) and postictal stertorous respirations (89%, p < 0.001). Only 10% of the 20 patients with asymmetric TC seizures had brief PGES; focal unaware seizures (N = 25) were not associated with PGES or stertorous breathing. Some patients (24%) with asymmetric or bilateral symmetric TC seizures had severe postictal encephalopathy with stertor that was separate or extended beyond periods of PGES.

CONCLUSION

Bilateral symmetric TC seizures, but not focal unaware seizures, have postictal stertor during PGES. Severe postictal encephalopathy, however, is also associated with motor dysfunction and stertor. Stertor appears to be a compensatory postictal respiratory pattern for ictal/postictal hypoxemia and occurs with PGES or postictal encephalopathy.

Let's talk SUDEP

Sudden unexplained death in epilepsy (SUDEP) is a devastating complication of epilepsy which was under-recognized in the recent past despite its clear importance. In this review, we examine the definition of SUDEP, revise current pathophysiological theories, discuss risk factors and preventative measures, disclose tools for appraising the SUDEP risk, and last but not least dwell upon announcing and explaining the SUDEP risk to the patients and their caretakers. We aim to aid the clinicians in their responsibility of knowing SUDEP, explaining the SUDEP risk to their patients in a reasonable and sensible way and whenever possible, preventing SUDEP. Future studies are definitely needed to increase scientific knowledge and awareness related to this prioritized topic with malign consequences.