Date of Completion


Embargo Period



Electrodermal Activity, Autonomic Nervous System, Sympathetic Nervous System, Heart Rate Variability, Carbon-Salt-Adhesive Electrodes

Major Advisor

Ki Chon

Associate Advisor

Kazunori Hoshino

Associate Advisor

Sabato Santaniello

Field of Study

Biomedical Engineering


Doctor of Philosophy

Open Access

Open Access


About 70 million people worldwide are affected by dysfunction of the autonomic nervous system (ANS). These dysfunctions can lead to life-threatening conditions, because ANS controls vital functions like heart rate, digestion, respiration, among others. Spectral analysis of heart rate variability (HRV) allows noninvasive assessment of the ANS. However, the low frequencies (0.04-0.15 Hz) of HRV are influenced by the two complementary branches of the ANS, the sympathetic and parasympathetic nervous systems; hence, the spectral analysis cannot separate the dynamics of these autonomic tones. Developing reliable noninvasive techniques for assessment of the ANS dynamics remains a challenge for scientists. Recently, measure of electrodermal activity (EDA) has gained popularity for ANS assessment. EDA signal represents the changes of conductance in the skin, produced by purely sympathetic innervation of sweat glands. Time-domain measures of EDA, such as the skin conductance level (SCL) and skin conductance responses (SCR), are correlated to the sympathetic control, but they lack of consistency. Surprisingly, EDA has been analyzed only in time domain. In this work, using a frequency domain measure via the spectral analysis, we established for the first time that frequency bands of the sympathetic tone in EDA is mostly confined to the 0.045-0.25 Hz range. We termed this power the EDASymp. Furthermore, utilizing a high-resolution time-varying approach we found that the frequency band of 0.08-0.24 Hz is most responsive to sympathetic tone, which we termed, the TVSymp. The EDASymp and TVSymp are more sensitive and consistent discriminators of orthostatic, physical and cognitive stress than time-domain measures of EDA and can quantitatively assess the sympathetic function. Moreover, we found that the sympathetic tone in EDA dynamically changes to higher frequencies when subjects exercise. This result can be useful for adjusting the bounds of HRV when heart rate increases. Currently, the same frequency bounds for the ANS for resting and exercise conditions have been used. Another application of EDA is monitoring of sleep deprivation. Using 24-hour sleep deprivation experiments, we found that high-frequencies (SCR) of EDA relate to vigilance whereas low-frequencies (SCL) relate to reactiveness. Based on these results, we found that EDA could potentially be used to assist in prevention of fatigue and stress which are the direct and dire consequences of prolonged wakefulness. As a final aim of my dissertation work, evaluations of novel dry carbon/salt/adhesive (CSA) electrodes for obtaining electrocardiogram, surface electromyogram and EDA were investigated. We found non-statistical difference between the carbon/salt/adhesive electrodes for the above mentioned applications when the electrodes were compared to the gold standard silver/silver chloride electrodes. Moreover, it was found that carbon electrodes exhibited better response to noise and motion artifacts when compared to hydrogel Ag/AgCl electrodes. Hence, we concluded that CSA electrodes can be reliably used as a surrogate of the Ag/AgCl electrodes along with the advantage of being more cost effective alternative as they have longer shelf-life.

Available for download on Wednesday, December 06, 2017