Form factor meets function: anatomy-dependent electrode-skin coupling and signal content in consumer eyewear EEG systems

Objective. To evaluate the influence of head morphology on the performance of a wearable setup that incorporates the constraints of an eyewear-electroencephalography (EEG) device suitable for consumer-level applications. Specifically, the study aimed to characterize the electrode-skin impedance of two dry-electrode types mounted on eyeglass frames, assess the system's ability to capture alpha-rhythm modulation during eyes-open and eyes-closed (EOEC) states in the temporal region, and its capability to detect auditory event-related potentials (P300).Approach. A prototype was built by embedding four EEG electrodes, two gold-plated retractile pins (GPR) and two conductive elastomer (CoE), into a commercial eyeglass frame, with reference and bias on the nose pads. Signals were acquired using an OpenBCI Cyton board (ADS1299 analog front end, sampling at 256 Hz). Twenty young healthy adults underwent three experimental protocols, namely electrode-skin contact assessment, EOEC tasks (two cycles of 2 min each) to examine alpha-band (8-12 Hz) power changes and compute an alpha-to-broadband power ratio, and an auditory oddball paradigm (80% standard, 20% odd stimuli, 50 odd trials) to elicit and analyze P300 components.Main results. GPR electrodes exhibited moderately higher median impedance but slightly narrower confidence intervals compared to CoE electrodes. Head breadth significantly affected GPR impedance (≈11.7Þcrease per mm increase), but had no significant effect on CoE impedance. Alpha-band power increased significantly during eyes-closed periods across subjects and electrode types. P300 responses (positive deflection at 300 ms) were reliably detected, with GPR electrodes yielding tighter latency distributions.Significance. These findings emphasize the importance of careful design considerations in wearable-EEG to account for inter-subject head anatomy variability and demonstrate that eyeglass-integrated EEG, can reliably capture both evoked and spontaneous neural responses.