Development of a continuous multisite accelerometry system for studying movements during sleep.

Actigraphy has proven to be a useful tool in the assessment of circadian rhythms, and more recently in the automatic staging of sleep and wake states. Whilst accuracy of commercial systems appears good over 24 hour periods, the sensitivity of detecting wake during time in bed is poor. One possible explanation for these poor results is the technical limitations of currently available commercial actigraphs. In particular, raw data is generally not available to the user. Instead, activity counts for each epoch (typically between 10-60 secs) are calculated using various algorithms, from which sleep state is identified. Consequently morphologically different movements observed during sleep and wake states may not be detected as such. In this paper, the development of a continuous multisite, accelerometry system (CMAS) is described. Initial results, comparing data collected using a commercial actigraph (Actiwatch- Mini Motionlogger), and the continuous multisite accelerometry system are presented. The CMAS is able to differentiate brief movement "twitches" from postural changes.