Development of an actigraph system for sleep-wake identification

Abstract

Serious sleep disorder may interfere with people's daily life, so monitoring the physiological signals and analyzing the recorded data are to find a solution to improve sleep quality. According to the survey of National Institute of Neurological Disorders and Stroke (NINDS) of United States, about 60 million Americans a year have insomnia frequently or for extended periods of time. A polysomnography (PSG) is a common approach with comprehensive multi-parametric physiological recording for sleep testing, but a subject sleeping in an unfamiliar environment with many wires attached will not feel comfortable and the recorded data cannot reflect the real sleeping quality. A wrist-watch actigraph recorder was developed for using within a living environment, while 35 healthy and 11 insomnia subjects were monitored and recorded by both PSG and actigraphy all night 8 hours for validation. The results show 91% in overall accuracy, 92% in sensitivity, 80% in specificity, 97% in predictive value for sleep and 54 % in predictive value for wake. © 2011 IEEE.