Abstracts

Rationale: Children with epilepsy, particularly those with refractory epilepsy, have significantly higher rates of adverse neurocognitive sequelae than age-matched cohorts. Early seizure detection via EEG is critical to both prevent adverse neurodevelopmental outcomes and to allow for urgent evaluation for an underlying seizure etiology. Current systems have numerous limitations such as cumbersome wired attachments that can impede clinical care and limits use outside of medical care facilities. We have developed a novel wearable 16-channel wireless EEG system that can overcome these limitations through 1) its compact design; 2) wireless output to a portable tablet; 3) advanced onboard electronics allowing children to move more freely with performance equivalent to significantly more expensive and cumbersome EEG systems.

Methods: A 16 channel EEG recording device was developed with specifications that meet or exceed current FDA-approved devices (1000 Hz sampling, 24 bit resolution, minimum signal amplitude 1uV). It is able to record 6.5 hours of continuous data and transmits it wirelessly to a standard mobile device via Bluetooth protocols. 14 patients, ages six months to 17 years (median age 9.5 years), were enrolled in an IRB approved prospective observational cohort study between January through May 2021 in the Lurie Epilepsy Monitoring Unit. 2-4 hours of EEG via the research EEG device was obtained from each patient concurrently with clinical standard EEG recording via Natus Xltek (Natus Medical, Pleasanton CA) via standard cup electrodes in 10-20 system positions. Data from both systems were standardized to the European Data Format (EDF) and viewed via EDFbrowser (Teunis van Beelen). Via this common interface, time-synchronized EEG studies from both the research system and Xltek were qualitatively reviewed by a pediatric epileptologist.

Results: Cerebral EEG signals were clearly visible for all 14 patients (100%) with the research system and clearly distinguishable from extra-cerebral signals such as eye blinking and jaw clenching. For every patient, the system displayed expected background features of wakefulness and sleep (posterior dominant rhythm, anterior-posterior gradient, sleep spindles). Eight patients had interictal epileptiform abnormalities (focal/multifocal spikes, generalized paroxysmal fast activity). Seizures were seen in two patients; one epileptic spasm, and one atypical absence seizure (Figure 1). All described EEG features were concordant between the research system and the gold-standard one. Upon qualitative review, there was intermittent low amplitude, high frequency noise present in the research system though it did not obscure any interictal or ictal abnormalities.

Conclusions: The novel wireless EEG system produced relevant background, interictal and ictal data that was qualitatively comparable to gold-standard recording, justifying future device development. Future directions include detailed quantitative comparison between the research and gold-standard recording system as well as the development of novel electrode solutions and automated signal analysis to increase the clinical applicability of the system.

Funding: Please list any funding that was received in support of this abstract.: Bill and Melinda Gates Foundation, INV-019423.