An Approach on Estimation of Sleep Cycle Using Occurrence Rate of Body Movements

Abstract

Accurate assessment of the sleep cycle at home will improve self-awareness regarding the necessity of sleep for relieving fatigue. Previous studies have shown that a large number of body movements occur during light sleep compared to deep sleep. In the present study, sleep was assessed by monitoring body movements using an infrared motion sensor and Polysomnography (PSG) for a 420-min period from 00:00 to 07:00. The findings of sleep cycle based on body movements were compared with PSG findings. One cycle of sleep cycle was defined as a set of “REM, sleep level 1, 2, 3 or 4” period and subsequent “SWS, sleep level 5 or 6” period. We defined Body Movement Density (BMD) as an index of sleep cycle. BMD was a value representing the number of body movement occurrences during a 30-minutes period. The cross-correlation coefficient between BMD and sleep levels was significant in all subjects (13 males, 3 females;age, 20-23 years). In this study, a concordance rate of 76.9% (SD±24.0) between BMD cycle and sleep cycle was obtained. This study indicates the possibility of estimation for sleep cycle by using BMD without using PSG. One of the greatest advantages of BMD is that it comprises only body movement data, so measurements of breath, heart rate or brain waves are not required. Moreover, measuring BDM is both inexpensive and noninvasive.