Technical Validation of the Zeto Wireless, Dry Electrode EEG System
Abstract number :
2.497
Submission category :
3. Neurophysiology / 3A. Video EEG Epilepsy-Monitoring
Year :
2023
Submission ID :
1386
Source :
www.aesnet.org
Presentation date :
12/3/2023 12:00:00 AM
Published date :
Authors :
Presenting Author: Zoltan Nadasdy, PhD – University of Texas at Austin
Robert Fisher, M.D., Ph.D. – Professor, Neurology & Neurological Sciences, Stanford University; Adam Fogarty, B.S. – EEG technologist, Neurology, Stanford University; Kevin Graber, M.D. – Clinical Professor, Neurology & Neurological Sciences, Stanford University; Christopher Primiani, M.D. – Adult Epilepsy Fellow, Neurology & Neurological Sciences, Stanford University
Rationale:
Wireless technology is transforming the practice of electroencephalography (EEG), but clinical adoption of the technology is contingent upon uncompromised data quality. Therefore, technical validation of the reliability of measurements made by a new device is critical, especially when multiple innovations are implemented. Zeto introduced a wireless dry electrode system, the FDA-approved WR-19, to improve patient comfort and save time. It provides easy application and removal and cloud-based remote monitoring. Approval required documentation of reliable signal quality. In this study, we replicate a favorable signal quality in the clinical setting on real patients.
Methods:
We simultaneously recorded 30-minute EEGs in thirteen sessions with a Nihon-Kohden JE -921A EEG system and Zeto’s full-montage WR-19 wireless headset using the Zeto dry electrodes and their cloud-based software. The wet and dry electrodes were placed adjacent to each other while keeping approximately compliant with the 10-20 positioning system. Subjects were sitting in a reclined chair with a back pillow or neck roll. The signals deriving from the two systems were converted to the same sampling rate (250 Hz) but the μV scale of amplitudes was kept. This allowed for a direct comparison of signal-to-signal, including signal correlation, spectral correlation, and signal-to-noise (SNR) estimations. >
Results:
Data were analyzed from 13 patients, ages between 19 and 70. Seven of them were females. Interpretations from conventional EEG included: abnormal (n=3), diffusely slow (n=1), focally slow (n=1), and none showing ictal activity. All 13 recordings resulted in acceptable quality EEG signals for 90% of the total recording duration on all 19 electrodes, and 10% were unreadable due to artifacts. Each of the 13 datasets showed an overall high spectral correlation (r >0.99; P< 0.001) across all electrode positions, indicating no systematic distortion of the signal at any frequency band. The signal-to-signal correlation was relatively high (r >0.6; P< 0.001), even though the two systems were not recorded at the exact same sampling frequency and electrodes were one to two cm apart. The dry electrode system's mean spectral SNR was 1 dB lower than the conventional system (P< 0.05) but with a 72% overlap between the two SNR distributions.
Neurophysiology