Abstracts

This is a Late Breaking abstract

Rationale: Electroencephalography (EEG) has long been the gold standard for the detection and characterization of seizures. However, the paroxysmal nature of seizures presents various barriers well recognized by clinicians. Routine EEG is limited by the rarity of events captured during brief recordings, and more prolonged ambulatory or inpatient EEG is often logistically impractical, technically challenging, expensive, and often still limited to only a few days of monitoring. As a result, clinicians often rely on subjective patient reports to assess disease burden and treatment efficacy, yet these have been shown to be inaccurate. Given these challenges, we hoped to leverage the comfort and wearability of a novel, small, portable, battery-powered Ear EEG system in a proof-of-concept comparative study. We compared the seizure detection performances of the NextSense Ear EEG system (ear-EEG) compared to traditional EEG. We also characterized the types of seizures and the anatomic boundaries of the events detected on the ear-EEG.

Methods: We recorded 1128 hours of simultaneous ear-EEG and scalp or intracranial EEG in 20 patients admitted to the Emory Epilepsy Monitoring Unit for presurgical or diagnostic evaluation. The ear-EEG records were first blindly reviewed and labeled for seizures by two epileptologists. The same reviewers subsequently annotated the scalp/stereotactic EEG records for comparative analysis.

Results: Results showed that 85.6% of all focal onset seizures (38.5 out of 45 episodes) were detected on the ear-EEG, with a false detection rate of 0.004 per hour (4 in 1128 hrs). The few episodes not detected on the ear-EEG emanated from deep within the mesial temporal lobe or extra-temporally and remained very focal. Results also showed that all focal onset seizures with secondary generalization (20 out of 20 episodes) were detected on the ear-EEG regardless of their originating anatomic lobe. All patients with seizures (13 out of 13) had at least one seizure detected on the ear-EEG.

Conclusions: These preliminary results demonstrate the feasibility and high accuracy of an ear-wearable device to record long-term EEG and capture seizures. We demonstrated that the NextSense Ear EEG system reliably captures temporal lobe seizures, which represent about one-third of all forms of epilepsy. In addition, 100% of seizures with secondary generalization were detected on ear-EEG, showing promise for capturing a large proportion of seizure types, including for patients with genetic generalized epilepsy, which characterize more than 40% of all epilepsies. The finding that all patients with seizures had at least one event detected on ear-EEG also suggests improved likelihood of detecting pathological events with prolonged monitoring capability. An unobtrusive yet reliable wearable device opens the door to routine collection of complementary, longitudinal, remote electrographic evidence that may assist clinicians in making an epilepsy diagnosis, assessing treatment efficacy, and optimizing medication titration.

Funding: Not applicable