Detection of focal to bilateral tonic-clonic seizures using a connected shirt.

OBJECTIVE: This study was undertaken to develop and evaluate a machine learning-based algorithm for the detection of focal to bilateral tonic-clonic seizures (FBTCS) using a novel multimodal connected shirt. METHODS: We prospectively recruited patients with epilepsy admitted to our epilepsy monitoring unit and asked them to wear the connected shirt while under simultaneous video-electroencephalographic monitoring. Electrocardiographic (ECG) and accelerometric (ACC) signals recorded with the connected shirt were used for the development of the seizure detection algorithm. First, we used a sliding window to extract linear and nonlinear features from both ECG and ACC signals. Then, we trained an extreme gradient boosting algorithm (XGBoost) to detect FBTCS according to seizure onset and offset annotated by three board-certified epileptologists. Finally, we applied a postprocessing step to regularize the classification output. A patientwise nested cross-validation was implemented to evaluate the performances in terms of sensitivity, false alarm rate (FAR), time in false warning (TiW), detection latency, and receiver operating characteristic area under the curve (ROC-AUC). RESULTS: We recorded 66 FBTCS from 42 patients who wore the connected shirt for a total of 8067 continuous hours. The XGBoost algorithm reached a sensitivity of 84.8% (56/66 seizures), with a median FAR of .55/24 h and a median TiW of 10 s/alarm. ROC-AUC was .90 (95% confidence interval = .88-.91). Median detection latency from the time of progression to the bilateral tonic-clonic phase was 25.5 s. SIGNIFICANCE: The novel connected shirt allowed accurate detection of FBTCS with a low false alarm rate in a hospital setting. Prospective studies in a residential setting with a real-time and online seizure detection algorithm are required to validate the performance and usability of this device.

Evaluation of Commercially Available Seizure Detection Wearables in Canada: Current Evidence.

Wearable-based seizure detection devices hold promise in reducing seizure-related adverse events and relieving the daily stress experienced by people with epilepsy. In this work, we present the latest evidence regarding the performance of three seizure detection wearables (eight studies) commercially available in Canada to provide guidance to clinicians. Overall, their ability to detect focal-to-bilateral and/or generalized tonic-clonic seizures ranges between 21.0% and 98.15% in sensitivity, with the 24h false alarm rates ranging from 0 to 1.28. While performance in epilepsy monitoring units show promise, the lack of evidence in outpatient settings precludes strong recommendations for their use in daily life.

Spatially Distinct Beta-Band Activities Reflect Implicit Sensorimotor Adaptation and Explicit Re-aiming Strategy.

Previous research suggests that so-called implicit and explicit processes of motor adaptation are implemented by distinct neural structures. Here we tested whether implicit sensorimotor adaptation and strategic re-aiming used to reduce movement error are reflected by spatially distinct EEG oscillatory components. We analyzed beta-band oscillations (∼13-30 Hz), which have long been linked to sensorimotor functions, at the time when these adaptive processes intervene for movement planning. We hypothesized that beta-band activity within sensorimotor regions relates to implicit adaptive processes, whereas beta-band activity within medial motor areas reflects deliberate re-aiming. In female and male human volunteers, we recorded EEG in a motor adaptation task in which a visual rotation was introduced in short series of trials separated by unperturbed trials. Participants were instructed in advance about the nature of the visual perturbation and trained to counter it by strategically re-aiming at a neighboring target. Consistent with our hypothesis, we found that preparatory beta-band activities within the two regions exhibited different patterns of modulation. Beta power in lateral central regions was attenuated when a change in the visual condition rendered internal-model predictions uncertain. In contrast, beta power in medial frontal regions was selectively decreased when participants strategically re-aimed their reaches. We propose that the reduction in lateral central beta power reflects an increased weighting of peripheral sensory information implicitly triggered when an adaptive change in the sensorimotor mapping is required, whereas the reduction in medial frontal beta-band activity relates to the inhibition of automatic motor responses in favor of cognitively controlled movements.SIGNIFICANCE STATEMENT Behavioral and modeling studies have proposed that so-called implicit and explicit components of motor adaptation recruit different neural circuits. Here, we investigated whether these different processes are reflected by spatially distinct beta-band activities. Analyzing EEG signals at the time they influence movement planning, during the foreperiod, we found that beta power within lateral central regions was decreased when a change in visual conditions required implicit sensorimotor remapping, which may reflect enhanced sensory processing when internal-model predictions are rendered uncertain. In contrast, beta-band power within medial frontal areas was selectively attenuated when participants deliberately re-aimed their movements to improve task performance, which may be associated with the inhibition of automatic motor responses in favor of cognitively controlled movements.

Error-related modulations of the sensorimotor post-movement and foreperiod beta-band activities arise from distinct neural substrates and do not reflect efferent signal processing.

While beta activity has been extensively studied in relation to voluntary movement, its role in sensorimotor adaptation remains largely uncertain. Recently, it has been shown that the post-movement beta rebound as well as beta activity during movement-preparation are modulated by movement errors. However, there are critical functional differences between pre- and post-movement beta activities. Here, we addressed two related open questions. Do the pre- and post-movement error-related modulations arise from distinct neural substrates? Do these modulations relate to efferent signals shaping muscle-activation patterns or do they reflect integration of sensory information, intervening upstream of the motor output? For this purpose, first we exploited independent component analysis (ICA) which revealed a double dissociation suggesting that distinct neural substrates are recruited in error-related beta-power modulations observed before and after movement. Second, we compared error-related beta oscillation responses observed in two bimanual reaching tasks involving similar movements but different interlimb coordination, and in which the same mechanical perturbations induced different behavioral adaptive responses. While the task difference was not reflected in the post-movement beta rebound, the pre-movement beta activity was differently modulated according to the interlimb coordination. Critically, we show an uncoupling between the behavioral and the electrophysiological responses during the movement preparation phase, which demonstrates that the error-related modulation of the foreperiod beta activity does not reflect changes in the motor output from primary motor cortex. It seems instead to relate to higher level processing of sensory afferents, essential for sensorimotor adaptation.