AAAI Publications, 2018 AAAI Spring Symposium Series

Font Size: 
Improving Sleep Stage Estimation Accuracy by Circadian Rhythm Extracted from a Low Frequency Component of Heart Rate
Akari Tobaru, Fumito Uwano, Takuya Iwase, Kazuma Matsumoto, Ryo Takano, Yusuke Tajima, Yuta Umenai, Keiki Takadama

Last modified: 2018-03-15


This paper described that proposing a novel method to estimate the sleep stage by biological data obtained with a non-contact sensor devices and that investigating its effectiveness. Proposed method focused on circadian rhythm to consider of a day biological rhythm in overall sleeping in addition to employ the Harada’s method. To verify the effectiveness of the proposed method, we derived the subject experiment that compared with the evaluation accuracy by the previous method with adjustment of circadian rhythm. As the experimental results, the following implications have been revealed: (1) the accuracy of the sleep stage estimation in 5 days out of that in 9 days were improved by proposed method in comparison with Harada’s method; (2) the parameter β (which determines the discount rate of curve of circadian rhythm) should be set around 60%, meaning that a raw circadian rhythm (i.e., no discounted rhythm) strongly affected the sleep stage while the highly discounted circadian rhythm (e.g. 30% discounted rhythm) does not contribute to accurately estimating the sleep stage.


sleep stage estimation; circadian rhythm; heart rate

Full Text: PDF