Correlating Motion Artifacts in Wet and Dry Electrodes With Head Kinematics During Physical Activities in Ambulatory EEG Monitoring

Electroencephalography (EEG) is a common neuroimaging technique used in clinical, research, and consumer technology given its noninvasiveness, high temporal resolution, and sensitivity. The main challenge in ambulatory EEG monitoring is the presence of substantial motion artifacts, especially for high-intensity movements such as running. This study compared motion artifacts in gold-cup, gel, and dry electrodes during different levels of physical activity. Eight healthy participants were instrumented with a wearable EEG cap and an instrumented mouthguard to measure head kinematics. While all electrode types performed similarly in the resting state, dry electrodes showed high-amplitude motion artifacts and worse performance during walking, running, and jumping. EEG-motion coherence was found to be higher for dry electrodes in low-intensity movements (e.g., walking with 0.1–0.5-g head acceleration, where 1 g =9.81 m/s
$^{2}$
), but similar across the three electrode types for medium-to-high-intensity movements (e.g., running and jumping with 0.5–3.5-g head acceleration), with peak coherences at up to 20 Hz. Fast walk trials showed the highest linear correlation between EEG signal and head acceleration, indicating potentially better coupling between the head and the electrode at this movement intensity. Our results imply that dry electrodes require further hardware development to mitigate motion artifacts before reliable application in ambulatory monitoring, and wet electrodes also exhibit clear movement-related artifacts in higher intensity movements. Uniquely, we identified EEG-motion coupling as a contributing factor to motion artifact characteristics, and coherence analysis may be a promising approach for motion artifact identification and quantification. Further characterization, quantification, and mitigation of motion artifacts will aid in higher quality measurement and meaningful interpretation of ambulatory EEG signals for clinical and research applications.