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.

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.

Objective: Wrist‐ or ankle‐worn devices are less intrusive than the widely used electroencephalographic (EEG) systems for monitoring epileptic seizures. Using custom‐developed deep‐learning seizure detection models, we demonstrate the detection of a broad range of seizure types by wearable signals.Methods: Patients admitted to the epilepsy monitoring unit were enrolled and asked to wear wearable sensors on either wrists or ankles. We collected patients' electrodermal activity, accelerometry (ACC), and photoplethysmography, from which blood volume pulse (BVP) is derived. Board‐certified epileptologists determined seizure onset, offset, and types using video and EEG recordings per the International League Against Epilepsy 2017 classification. We applied three neural network models—a convolutional neural network (CNN) and a CNN–long short‐term memory (LSTM)‐based generalized detection model and an autoencoder‐based personalized detection model—to the raw time‐series sensor data to detect seizures and utilized performance measures, including sensitivity, false positive rate (the number of false alarms divided by the total number of nonseizure segments), number of false alarms per day, and detection delay. We applied a 10‐fold patientwise cross‐validation scheme to the multisignal biosensor data and evaluated model performance on 28 seizure types.Results: We analyzed 166 patients (47.6% female, median age = 10.0 years) and 900 seizures (13 254 h of sensor data) for 28 seizure types. With a CNN‐LSTM‐based seizure detection model, ACC, BVP, and their fusion performed better than chance; ACC and BVP data fusion reached the best detection performance of 83.9% sensitivity and 35.3% false positive rate. Nineteen of 28 seizure types could be detected by at least one data modality with area under receiver operating characteristic curve > .8 performance.Significance: Results from this in‐hospital study contribute to a paradigm shift in epilepsy care that entails noninvasive seizure detection, provides time‐sensitive and accurate data on additional clinical seizure types, and proposes a novel combination of an out‐of‐the‐box monitoring algorithm with an individualized person‐oriented seizure detection approach.

Objective: Daytime and nighttime patterns affect the dynamic modulation of brain and body functions and influence the autonomic nervous system response to seizures. Therefore, we aimed to evaluate 24‐hour patterns of electrodermal activity (EDA) in patients with and without seizures.Methods: We included pediatric patients with (a) seizures (SZ), including focal impaired awareness seizures (FIAS) or generalized tonic‐clonic seizures (GTCS), (b) no seizures and normal electroencephalography (NEEG), or (c) no seizures but epileptiform activity in the EEG (EA) during vEEG monitoring. Patients wore a device that continuously recorded EDA and temperature (TEMP). EDA levels, EDA spectral power, and TEMP levels were analyzed. To investigate 24‐hour patterns, we performed a nonlinear mixed‐effects model analysis. Relative mean pre‐ictal (−30 min to seizure onset) and post‐ictal (I: 30 min after seizure offset; II: 30 to 60 min after seizure offset) values were compared for SZ subgroups.Results: We included 119 patients (40 SZ, 17 NEEG, 62 EA). EDA level and power group‐specific models (SZ, NEEG, EA) (h = 1; P < .01) were superior to the all‐patient cohort model. Fifty‐nine seizures were analyzed. Pre‐ictal EDA values were lower than respective 24‐hour modulated SZ group values. Post hoc comparisons following the period‐by‐seizure type interaction (EDA level: = 18.50; P < .001, and power: = 6.73; P = .035) revealed that EDA levels were higher in the post‐ictal period I for FIAS and GTCS and in post‐ictal period II for GTCS only compared to the pre‐ictal period.Significance: Continuously monitored EDA shows a pattern of change over 24 hours. Curve amplitudes in patients with recorded seizures were lower as compared to patients who did not exhibit seizures during the recording period. Sympathetic skin responses were greater and more prolonged in GTCS compared to FIAS. EDA recordings from wearable devices offer a noninvasive tool to continuously monitor sympathetic activity with potential applications for seizure detection, prediction, and potentially sudden unexpected death in epilepsy (SUDEP) risk estimation.

Physical exercise has been shown to modulate activity within the autonomic nervous system (ANS). Considering physical exercise as a holistic stimulus on the nervous system and specifically the ANS, uni- and multimodal analysis tools were applied to characterize centrally driven interactions and control of ANS functions. Nineteen young and physically active participants performed treadmill tests at individually determined moderate and high intensities. Continuous electrodermal activity (EDA), heart rate (HR), and skin temperature at wrist (Temp) were recorded by wireless multisensor devices (Empatica^® E4, Milan, Italy) before and 30 min after exercise. Artifact-free continuous 3 min intervals were analyzed. For unimodal analysis, mean values were calculated, for bimodal and multimodal analysis canonical correlation analysis (CCA) was performed. Unimodal results indicate that physical exercise affects ANS activity. More specifically, Temp increased due to physical activity (moderate intensity: from 34.15∘C to 35.34∘C and high intensity: from 34.11∘C to 35.09∘C). HR increased more for the high (from 60.76 bpm to 79.89 bpm) than for the moderate (from 64.81 bpm to 70.83 bpm) intensity. EDA was higher for the high (pre: 8.06 μS and post: 9.37 μS) than for the moderate (pre: 4.31 μS and post: 3.91 μS) intensity. Bimodal analyses revealed high variations in correlations before exercise. The overall correlation coefficient showed varying correlations in pretest measures for all modality pairs (EDA-HR, HR-Temp, Temp-EDA at moderate: 0.831, 0.998, 0.921 and high: 0.706, 0, 0.578). After exercising at moderate intensity coefficients changed little (0.828, 0.744, 0.994), but increased substantially for all modality pairs after exercising at high intensity (0.976, 0.898, 0.926). Multimodal analysis confirmed bimodal results. Exercise-induced changes in ANS activity can be found in multiple ANS modalities as well as in their interactions. Those changes are intensity-specific: with higher intensity the interactions increase. Canonical correlations between different ANS modalities may therefore offer a feasible approach to determine exercise induced modulations of ANS activity.