Toward a taxonomy of autonomic sleep patterns with electrodermal activity

Abstract

This paper presents a first version of a taxonomy of automatic sleep patterns found with the Affectiva Q™ Sensor, a wireless, logging biosensor that measures skin conductance, skin temperature, and motion comfortably from the wrist. Several studies have examined electrodermal activity (EDA) during sleep, but they focused on an analysis of EDA for only a small number of nights. We quantitatively analyzed EDA during sleep in three study situations: (1) Comparing EDA with polysomnography (PSG) from seven subjects in a sleep lab, (2) Characterizing multiple nights of EDA in a sleep lab, in a hospital and at home from 24 subjects, and (3) Gathering long-term EDA (30–60 nights) patterns from three subjects during home sleep. After gathering this rich corpus of data, we characterized inter- and intra-individual differences of EDA features and the relation of EDA peaks to subjective sleep quality. Here we present results from the three studies in an effort to begin to characterize autonomic patterns found in natural sleep.