Intra- and Inter-Subject Perspectives on the Detection of Focal Onset Motor Seizures in Epilepsy Patients

Feasibility of cardiac-based seizure detection and prediction: A systematic review of non-invasive wearable sensor-based studies

A reliable seizure detection or prediction device can potentially reduce the morbidity and mortality associated with epileptic seizures. Previous findings indicating alterations in cardiac activity during seizures suggest the usefulness of cardiac parameters for seizure detection or prediction. This study aims to examine available studies on seizure detection and prediction based on cardiac parameters using non-invasive wearable devices. The Embase, PubMed, and Scopus databases were used to systematically search according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines. Human studies that evaluated seizure detection or prediction based on cardiac parameters collected using wearable devices were included. The QUADAS-2 tool and proposed standards for validation for seizure detection devices were used for quality assessment. Twenty-four articles were identified and included in the analysis. Twenty studies evaluated seizure detection algorithms, and four studies focused on seizure prediction. Most studies used either a wrist-worn or chest-worn device for data acquisition. Among the seizure detection studies, cardiac parameters utilized for the algorithms mainly included heart rate (HR) (n = 11) or a combination of HR and heart rate variability (HRV) (n = 6). HR-based seizure detection studies collectively reported a sensitivity range of 56%-100% and a false alarm rate (FAR) of 0.02-8/h, with most studies performing retrospective validation of the algorithms. Three of the seizure prediction studies retrospectively validated multimodal algorithms, combining cardiac features with other physiological signals. Only one study prospectively validated their seizure prediction algorithm using HRV extracted from ECG data collected from a custom wearable device. These studies have demonstrated the feasibility of using cardiac parameters for seizure detection and prediction with wearable devices, with varying algorithmic performance. Many studies are in the proof-of-principle stage, and evidence for real-time detection or prediction is currently limited. Future studies should prioritize further refinement of the algorithm performance with prospective validation using large-scale longitudinal data. PLAIN LANGUAGE SUMMARY: This systematic review highlights the potential use of wearable devices, like wristbands, for detecting and predicting seizures via the measurement of heart activity. By reviewing 24 articles, it was found that most studies focused on using heart rate and changes in heart rate for seizure detection. There was a lack of studies looking at seizure prediction. The results were promising but most studies were not conducted in real-time. Therefore, more real-time studies are needed to verify the usage of heart activity-related wearable devices to detect seizures and even predict them, which will be beneficial to people with epilepsy.

Data quality evaluation in wearable monitoring

Wearable recordings of neurophysiological signals captured from the wrist offer enormous potential for seizure monitoring. Yet, data quality remains one of the most challenging factors that impact data reliability. We suggest a combined data quality assessment tool for the evaluation of multimodal wearable data. We analyzed data from patients with epilepsy from four epilepsy centers. Patients wore wristbands recording accelerometry, electrodermal activity, blood volume pulse, and skin temperature. We calculated data completeness and assessed the time the device was worn (on-body), and modality-specific signal quality scores. We included 37,166 h from 632 patients in the inpatient and 90,776 h from 39 patients in the outpatient setting. All modalities were affected by artifacts. Data loss was higher when using data streaming (up to 49% among inpatient cohorts, averaged across respective recordings) as compared to onboard device recording and storage (up to 9%). On-body scores, estimating the percentage of time a device was worn on the body, were consistently high across cohorts (more than 80%). Signal quality of some modalities, based on established indices, was higher at night than during the day. A uniformly reported data quality and multimodal signal quality index is feasible, makes study results more comparable, and contributes to the development of devices and evaluation routines necessary for seizure monitoring.

Seizure detection based on wearable devices: A review of device, mechanism, and algorithm

With sudden and unpredictable nature, seizures lead to great risk of the secondary damage, status epilepticus, and sudden unexpected death in epilepsy. Thus, it is essential to use a wearable device to detect seizure and inform patients' caregivers for assistant to prevent or relieve adverse consequence. In this review, we gave an account of the current state of the field of seizure detection based on wearable devices from three parts: devices, physiological activities, and algorithms. Firstly, seizure monitoring devices available in the market primarily involve wristband-type devices, patch-type devices, and armband-type devices, which are able to detect motor seizures, focal autonomic seizures, or absence seizures. Secondly, seizure-related physiological activities involve the discharge of brain neurons presented, autonomous nervous activities, and motor. Plenty of studies focus on features from one signal, while it is a lack of evidences about the change of signal coupling along with seizures. Thirdly, the seizure detection algorithms developed from simple threshold method to complicated machine learning and deep learning, aiming at distinguish seizures from normal events. After understanding of some preliminary studies, we will propose our own thought for future development in this field.

EEG Signals Based Internet Addiction Diagnosis Using Convolutional Neural Networks

Internet addiction (IA), as a new and often unrecognized psychosocial disorder, endangers people’s health and their lives. However, the common biometric analysis based on the combination of EEG signals and results of questionnaires is not quantitative, and thus difficult to ensure a specific biomarker. This work aims to develop a deep learning algorithm (no need to identify biomarkers) used for diagnosing IA and evaluating therapy efficacy. Herein, a five-layer CNN model combined with a fast Fourier transform is proposed to diagnose IA quantitatively. This algorithm is validated in the Lemon dataset by using it to process raw data, full spectral power, and alpha-beta-gamma spectral power (related to IA). In contrast to alpha-beta-gamma spectral power, the results based on full spectral power show better performance (87.59% accuracy, 88.80% sensitivity, and 86.41% specificity), which confirms that the proposed algorithm can diagnose IA without biomarkers. In addition, this proposed CNN model presents obvious advantages in processing raw data, achieving 81.1% accuracy. Such results verify that this method can contribute to the reduction of diagnosis time and be potentially used in real-time health monitoring systems. This work provides a quantitative approach to diagnose IA and evaluate therapy efficacy, as a general strategy, and can be widely used in other disorder diagnoses that affect EEG signals, such as psychiatric disorders, substance dependence, and depression.