Profiling continuous sleep representations for better understanding of the dynamic character of normal sleep

Highlights：

• The sleep process is represented by continuous sleep probabilistic curves.

• The method iteratively combines cluster analysis and time alignment of sleep probabilistic curves.

• The sleep profiles associated with individual clusters help to better understand existing associations between the dynamic of sleep states and daily measures.

• The results are consistent with existing studies, which makes the proposed methodology a promising tool in sleep research.

摘要

•A novel framework for identification of specific continuous sleep profiles significantly associated with a set of variables representing daytime subjective, neurophysiological and cognitive states of a population is proposed.•The sleep process is represented by continuous sleep probabilistic curves.•The method iteratively combines cluster analysis and time alignment of sleep probabilistic curves.•The sleep profiles associated with individual clusters help to better understand existing associations between the dynamic of sleep states and daily measures.•The results are consistent with existing studies, which makes the proposed methodology a promising tool in sleep research.

论文关键词：Sleep probabilistic model,Sleep probabilistic curves,Functional cluster analysis,Time alignment,Dynamic time warping,Daily life measures

论文评审过程：Received 2 July 2018, Revised 12 December 2018, Accepted 27 December 2018, Available online 29 December 2018, Version of Record 13 June 2019.