Observer based on body movement information in sleeping and estimation of sleep stage appearance probability

Yosuke Kurihara, Kajiro Watanabe, Kazuyuki Kobayashi, Hiroshi Tanaka

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)

Abstract

The manual for scoring sleep defined by the American Academy of Sleep Medicine in 2007 contains some rules that, even as an international standard of sleep stage judgment, include ambiguities and are therefore compensated by subjective interpretations of sleep stage scorers. This paper presents a novel method for compensating the subjective interpretations and judgments and describing the judgments in probabilistic terms. We employed a full-order Luenberger observer (state estimation method) based on two models of sleep transition: no body movement and body movement. Sleep stages judged by three different scorers under the rules of the manual were rejudged by the observer. The average values of κ statistics, which show the degree of agreement, were 0.83, 0.89 and 0.81, respectively, for the original sleep stages. Because the new method provides probabilities on how surely the sleep belongs to each sleep stage, we were able to determine the most probable, the second most probable and the third most probable sleep stages. The K statistics between the most probable sleep stages were improved to 0.89, 0.93 and 0.85, respectively. Those of sleep stages determined from the most and second most probable were 0.93, 0.96 and 0.90 and those from the most, second most and third most probable were 0.95, 0.97 and 0.92.

Original languageEnglish
Pages (from-to)688-695
Number of pages8
JournalIEEJ Transactions on Electrical and Electronic Engineering
Volume3
Issue number6
DOIs
Publication statusPublished - 2008 Nov

Keywords

  • Sleep stage
  • Sleep stage state variable equation
  • Sleep stage transition probability matrix

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Observer based on body movement information in sleeping and estimation of sleep stage appearance probability'. Together they form a unique fingerprint.

Cite this