Poster K16 - Towards real-world application of blink-related brainwaves for brain function assessment: Extracting blink-related oscillations using low-density electroencephalography with dry and textile electrodes
Introduction: Blink-related oscillations (BROs) are brainwave responses that occur following spontaneous blinking and correspond to neural mechanisms as the brain processes the new image that appear after the eyes reopen. These responses have been shown to be useful for evaluating brain function in a variety of situations, and have potential as a digital biomarker for monitoring cognition. However, previous BRO studies have all employed large, fixed-infrastructure hardware systems such as magnetoencephalography (MEG) and high-density electroencephalography (EEG) with 64 channels, which are restricted to research settings and cannot be applied in everyday life. In order to create a BRO-based technology for daily cognition monitoring, it is essential to develop hardware systems that are portable, easy to use, and comfortable to wear. Towards this end, we undertook the first investigation of BRO measurements using data from a portable, low-density EEG system with only 4 channels and textile-based electrodes for faster application and enhanced participant comfort.
Materials and
Methods: This study utilized existing, publicly available data from Lopez et al (2023), containing 3 minutes of EEG data collected in 10 healthy adults under eyes-open resting conditions. The signals were measured using both textile-based electrodes and dry Ag/Ag-Cl electrodes at the same 4 electrode locations (Fp1, Fp2, F7, F8) and using the same EEG amplifier hardware. The signals were denoised using an established time-frequency filtering approach, then transformed back into the time domain to derive BRO responses. BRO measurements were compared between the textile and dry Ag/Ag-Cl electrodes to examine effects of artifact removal and signal capture in both time and frequency domains, and morphological comparisons were made using intraclass correlation. Statistical comparison was undertaken using paired t-test and Wilcoxon signed rank test.
Results, Conclusions, and Discussions: BRO responses measured using both textile and dry electrodes exhibited similar morphologies in both time and frequency domains (Figure 1a,b), consistent with prior literature. BRO response amplitudes were also comparable between the textile and dry electrodes in both the post-blink peaks as well as the pre-blink baseline (Figure 1c). Both sensor technologies showed similar extent of ocular artifact removal following signal denoising, with >99% reduction in ocular signal contribution. Additionally, there was high intraclass correlation between the two sensor technologies in both time and frequency domains, suggesting similar morphology in BRO responses that were captured between the textile and dry Ag/Ag-Cl electrodes. Moreover, individual-level BRO response amplitudes and latencies were also highly correlated between the two sensor technologies, suggesting that the measurements were reliable at the individual level. Together, these results demonstrate for the first time that BRO responses can be successfully captured using a low-density EEG system with textile-based electrodes and only 4 channels, with comparable performance compared to dry Ag/Ag-Cl electrodes. These results provide compelling evidence towards the development of a future BRO-based technology that is fast and easy to apply, comfortable to wear, and can provide objective measurements of brain function that leverage the simplicity and ubiquity of spontaneous blinking to maximize application potential in everyday life.
