REM /NREM Discrimination via Ocular and Limb Movement Monitoring: Correlation with Polygraphic Data and Development of a REM State Algorithm.

The purpose of this study was to develop an inexpensive, simplified home monitoring system for prescreening certain sleep complaints and for long-term home-based experiments. Four subjects slept in the laboratory for 3-4 nights. In addition to a standard polysomnogram, piezoelectric ceramic transducers were attached to both hands, both feet, and over the right eyelid. The sleep from 2 nights of each subject was staged according to the standard criteria. The limb transducers identified major and minor body shifts and were used to signal possible state changes. The data from the eye movement sensor was used to develop a REM-state algorithm. Using this algorithm, REM and NREM sleep were identified with 92% (SD=0.091) and 91% (SD=0.076) accuracy, respectively. The ultimate goal is to have a bedside device programmed with the algorithm that would either store information or score the data on-line for waking, NREM, and REM states. We conclude that a system monitoring eye movements and major body shifts is a promising approach to an economical and accurate home recording system. [ABSTRACT FROM AUTHOR]