Background: With a wide range of use cases in both research and clinical domains, collecting continuous mobile health (mHealth) streaming data from multiple sources in a secure, highly scalable, and extensible platform is of high interest to the open source mHealth community. The European Union Innovative Medicines Initiative Remote Assessment of Disease and Relapse-Central Nervous System (RADAR-CNS) program is an exemplary project with the requirements to support the collection of high-resolution data at scale; as such, the Remote Assessment of Disease and Relapse (RADAR)-base platform is designed to meet these needs and additionally facilitate a new generation of mHealth projects in this nascent field. Objective: Wide-bandwidth networks, smartphone penetrance, and wearable sensors offer new possibilities for collecting near-real-time high-resolution datasets from large numbers of participants. The aim of this study was to build a platform that would cater for large-scale data collection for remote monitoring initiatives. Key criteria are around scalability, extensibility, security, and privacy. Methods: RADAR-base is developed as a modular application; the backend is built on a backbone of the highly successful Confluent/Apache Kafka framework for streaming data. To facilitate scaling and ease of deployment, we use Docker containers to package the components of the platform. RADAR-base provides 2 main mobile apps for data collection, a Passive App and an Active App. Other third-Party Apps and sensors are easily integrated into the platform. Management user interfaces to support data collection and enrolment are also provided. Results: General principles of the platform components and design of RADAR-base are presented here, with examples of the types of data currently being collected from devices used in RADAR-CNS projects: Multiple Sclerosis, Epilepsy, and Depression cohorts. Conclusions: RADAR-base is a fully functional, remote data collection platform built around Confluent/Apache Kafka and provides off-the-shelf components for projects interested in collecting mHealth datasets at scale.

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.

Background: Video electroencephalography recordings, routinely used in epilepsy monitoring units, are the gold standard for monitoring epileptic seizures. However, monitoring is also needed in the day-to-day lives of people with epilepsy, where video electroencephalography is not feasible. Wearables could fill this gap by providing patients with an accurate log of their seizures. Objective: Although there are already systems available that provide promising results for the detection of tonic-clonic seizures (TCSs), research in this area is often limited to detection from 1 biosignal modality or only during the night when the patient is in bed. The aim of this study is to provide evidence that supervised machine learning can detect TCSs from multimodal data in a new data set during daytime and nighttime. Methods: An extensive data set of biosignals from a multimodal watch worn by people with epilepsy was recorded during their stay in the epilepsy monitoring unit at 2 European clinical sites. From a larger data set of 243 enrolled participants, those who had data recorded during TCSs were selected, amounting to 10 participants with 21 TCSs. Accelerometry and electrodermal activity recorded by the wearable device were used for analysis, and seizure manifestation was annotated in detail by clinical experts. Ten accelerometry and 3 electrodermal activity features were calculated for sliding windows of variable size across the data. A gradient tree boosting algorithm was used for seizure detection, and the optimal parameter combination was determined in a leave-one-participant-out cross-validation on a training set of 10 seizures from 8 participants. The model was then evaluated on an out-of-sample test set of 11 seizures from the remaining 2 participants. To assess specificity, we additionally analyzed data from up to 29 participants without TCSs during the model evaluation. Results: In the leave-one-participant-out cross-validation, the model optimized for sensitivity could detect all 10 seizures with a false alarm rate of 0.46 per day in 17.3 days of data. In a test set of 11 out-of-sample TCSs, amounting to 8.3 days of data, the model could detect 10 seizures and produced no false positives. Increasing the test set to include data from 28 more participants without additional TCSs resulted in a false alarm rate of 0.19 per day in 78 days of wearable data. Conclusions: We show that a gradient tree boosting machine can robustly detect TCSs from multimodal wearable data in an original data set and that even with very limited training data, supervised machine learning can achieve a high sensitivity and low false-positive rate. This methodology may offer a promising way to approach wearable-based nonconvulsive seizure detection.

Objectives: High counts of averaged interictal epileptiform discharges (IEDs) are key components of accurate interictal electric source imaging (ESI) in patients with focal epilepsy. Automated detections may be time‐efficient, but they need to identify the correct IED types. Thus we compared semiautomated and automated detection of IED types in long‐term video‐EEG (electroencephalography) monitoring (LTM) using an extended scalp EEG array and short‐term high‐density EEG (hdEEG) with visual detection of IED types and the seizure‐onset zone (SOZ).Methods: We prospectively recruited consecutive patients from four epilepsy centers who underwent both LTM with 40‐electrode scalp EEG and short‐term hdEEG with 256 electrodes. Only patients with a single circumscribed SOZ in LTM were included. In LTM and hdEEG, IED types were identified visually, semiautomatically and automatically. Concordances of semiautomated and automated detections in LTM and hdEEG, as well as visual detections in hdEEG, were compared against visually detected IED types and the SOZ in LTM.Results: Fifty‐two of 62 patients with LTM and hdEEG were included. The most frequent IED types per patient, detected semiautomatically and automatically in LTM and visually in hdEEG, were significantly concordant with the most frequently visually identified IED type in LTM and the SOZ. Semiautomated and automated detections of IED types in hdEEG were significantly concordant with visually identified IED types in LTM, only when IED types with more than 50 detected single IEDs were selected. The threshold of 50 detected IED in hdEEG was reached in half of the patients. For all IED types per patient, agreement between visual and semiautomated detections in LTM was high.Significance: Semiautomated and automated detections of IED types in LTM show significant agreement with visually detected IED types and the SOZ. In short‐term hdEEG, semiautomated detections of IED types are concordant with visually detected IED types and the SOZ in LTM if high IED counts were detected.