Electrodermal Activity Monitoring With Flexible Dry Electrodes and AC Measurement

Wearable sensing technology has transformed the healthcare landscape, shifting from traditional hospital settings to more personalized home-based care. One prominent focus lies in evaluating autonomic nervous system activity, which has been made possible by integrating electrodermal activity (EDA) monitoring into wearable devices. In this study, we thoroughly explored measurement methods and electrode designs to identify the optimal solution for accurate EDA monitoring. The state-of-the-art method uses dc for measurement, which fails to capture the skin’s capacitive properties. Therefore, we explore using ac with flexible dry electrodes and examine the influence of electrode geometry (including gap size), material, and anatomical locations on biosignal quality. Custom dry electrodes with interdigitated design and gold plates exhibited promising results, offering lower impedance, shorter stabilization periods, and higher correlation with the standard pregelled silver-silver chloride (Ag/AgCl) electrodes. We found electrode-skin contact area critical in determining the signal quality for dry electrodes. We further compared the performance of the ac method across various designs using both handcrafted and deep learning (DL) extracted features. The results demonstrated that the constant-amplitude, ac-voltage exosomatic method significantly enhances EDA measurements. Specifically, for the application of cognitive stress detection, using phase angle information improved the accuracy by 12.72% compared with the dc method, resulting in an overall accuracy of 83.54% among ten participants. Thus, using an ac-based approach with an optimized dry electrode design could serve as a viable and accurate alternative to wet electrodes for long-duration monitoring.