This study aims at sleep stage classification problem via EOG signals.
The classification problem consists of four steps. The first step is to
distinguish slow wave sleep from the rest of the sleep periods. Wake periods
are identified in the second step. The third step finds REM sleep and the last
step classifies stage 2 and stage1 sleep.
By using different EOG signal features in different steps of the
classification process, this work uses back-propagation trained neural
networks to perform classification.
With the exception of stage 1 sleep, the sensitivity and positive
predictive value ranges from 70% to 80%. The overall classification accuracy
is 74.80%.
Identifer | oai:union.ndltd.org:NSYSU/oai:NSYSU:etd-0715109-205535 |
Date | 15 July 2009 |
Creators | Chen, Tao-hsin |
Contributors | none, Chen-Wen Yen, none |
Publisher | NSYSU |
Source Sets | NSYSU Electronic Thesis and Dissertation Archive |
Language | Cholon |
Detected Language | English |
Type | text |
Format | application/pdf |
Source | http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0715109-205535 |
Rights | unrestricted, Copyright information available at source archive |
Page generated in 0.0021 seconds