Experimental Evaluation of an Adaptive Gain Sensor for Electrodermal Activity Monitoring in a Smart Sock Form Factor

Wearable devices with Electrodermal Activity (EDA) sensors have been proposed for a wide range of applications such as affective computing. E-textiles are at the forefront of the development of comfortable and unobtrusive form factors, and smart socks offer an alternative form factor with enhanced nervous system monitoring capabilities compared to wrist-worn devices. Nonetheless, open problems remain unsolved such as scalability and problems with signal saturation due to the unique skin characteristics of the foot. In this paper, a smart sock was manufactured through industrially scalable methods, and an EDA sensor with an improved measurement range was used to overcome these issues. The prototype was tested in 15 subjects simultaneously with a reference device using pre-gelled electrodes at the hand. Pearson correlations of up to 95% were obtained for EDA signals, with no observed cases of signal saturation. Moreover, the number of co-occurring onsets and signal latency between hand and foot-EDA were similar to studies in which pre-gelled electrodes were used at both the hand and the foot. The improved EDA sensor resulted in minimal artifacts induced by its adaptive mechanism. Overall, this work supports existing evidence of the feasibility of using smart socks as a form factor for EDA monitoring.