Huupponen Eero, Himanen Sari-Leena, Hasan Joel, Värri Alpo
Signal Processing Laboratory, Tampere University of Technology, Korkeakoulunkatu 1, FIN 33101 Tampere, Finland.
J Med Syst. 2003 Aug;27(4):337-45. doi: 10.1023/a:1023753203633.
The automatic sleep analysis aims at providing an accurate description of sleep process. We found that there exist so far poorly known sleep depth oscillations constantly. Quantitative analysis of these oscillations was done in this work via a mean frequency measure and FFT. Overall charasteristics of these oscillations were studied, focusing on the waves with period times of 5-150 s. These sleep depth oscillations have a relatively large amplitude and they should be considered in future sleep analysis systems. The results of this study give directions to automated sleep analysis regarding optimal estimation of sleep depth.
自动睡眠分析旨在准确描述睡眠过程。我们发现,目前一直存在着鲜为人知的睡眠深度振荡。在这项工作中,通过平均频率测量和快速傅里叶变换对这些振荡进行了定量分析。研究了这些振荡的总体特征,重点关注周期为5 - 150秒的波。这些睡眠深度振荡具有相对较大的幅度,在未来的睡眠分析系统中应予以考虑。本研究结果为自动睡眠分析中睡眠深度的最佳估计提供了指导。