Abstracts

Rationale: Common concerns for people with generalized tonic-clonic seizures (GTCS), and their caregivers, include lack of independence and peace of mind that a GTCS will be detected should one occur. Given the diverse motor manifestations of seizures, additional routine motor characterization of seizures would have a positive impact on epilepsy management. Currently there are no FDA-cleared devices outside of an epilepsy monitoring unit (EMU) that can reliably alert for GTCS or provide accurate information about semiology to their physicians. The aim of this study is to validate the effectiveness of a novel EMG-based GTCS detection system that can be discreetly worn without interfering with activities of daily living. Continuous EMG monitoring would help patients with GTCS regain independence and provide accurate biological data for their epilepsy management. Methods: In this phase III double-blind controlled trial, we are testing the ability of the Brain Sentinel system to detect GTCS as compared to vEEG detection in epilepsy monitoring units (EMU) at 11 NAEC level IV Epilepsy Centers in the United States. The EMG data is processed in real-time by an algorithm previously validated in a single-site trial (Szabo et al., 2013). The classic GTCS semiology and timeline were independently adjudicated for all vEEG events recorded by a panel of ABPN certified epileptologists. The real time alert status of Brain Sentinel system was classified as true positives, false positives, true negatives and false negatives based upon vEEG. If GTCS occurred in EMU, subjects took the Brain Sentinel system home for two weeks. Results: To date, 1598 hours of EMG and vEEG data have been fully reviewed from 26 subjects admitted to the EMUs. Provisional analysis reveals sensitivity detection for GTCS that is approaching the sensitivity of current FDA-cleared automated EEG seizure detection algorithms. Following Brain Sentinel system detections of GTCS, review of the EMG data file revealed several components of motor manifestations, such as duration and temporal separation of tonic and clonic activity that help the semiological description of GTCS. Final device performance will be presented for the full study cohort (over 120 subjects) at the time of the conference. Conclusions: The sensitivity and predictive value of the alerts triggered by the Brain Sentinel system were compared to the gold standard of vEEG. The Brain Sentinel system allowed people with epilepsy to carryout daily activities with greater independence and peace of mind in the home. The device was also helpful in the semiological classification of motor manifestations during vEEG and an aid for the interpretation of home events.