Automatic REM Sleep Detection Associated with Idiopathic REM Sleep Behavior Disorder

Jacob Kempfner, Gertrud Laura Sørensen, Helge Bjarup Dissing Sørensen, P Jennum

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Rapid eye movement sleep Behavior Disorder (RBD) is a strong early marker of later development of Parkinsonism. Currently there are no objective methods to identify and discriminate abnormal from normal motor activity during REM sleep. Therefore, a REM sleep detection without the use of chin electromyography (EMG) is useful. This is addressed by analyzing the classification performance when implementing two automatic REM sleep detectors. The first detector uses the electroencephalography (EEG), electrooculography (EOG) and EMG to detect REM sleep, while the second detector only uses the EEG and EOG. Method: Ten normal controls and ten age matched patients diagnosed with RBD were enrolled. All subjects underwent one polysomnographic (PSG) recording, which was manual scored according to the new sleep-scoring standard from the American Academy of Sleep Medicine. Based on the manual scoring, an automatic computerized REM detection algorithm has been implemented, using wavelet packet combined with artificial neural network. Results: When using the EEG, EOG and EMG modalities, it was possible to correctly classify REM sleep with an average Area Under Curve (AUC) equal to 0:900:03 for normal subjects and AUC = 0:810:05 for RBD subjects. The performance difference between the two groups was significant (p0:05) in performance was observed when only using the EEG and EOG in neither of the groups. Conclusion: The overall result indicates that the EMG does not play an important role when classifying REM sleep.
Original languageEnglish
Title of host publicationIEEE Engineering in medicine and biology society conference proceedings
PublisherIEEE
Publication date2011
Pages6063-6066
ISBN (Print)978-1-4244-4122-8
DOIs
Publication statusPublished - 2011
Event33rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society - Boston, Massachusetts, United States
Duration: 30 Aug 20113 Sep 2011
Conference number: 33
http://embc2011.embs.org/

Conference

Conference33rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society
Number33
CountryUnited States
CityBoston, Massachusetts
Period30/08/201103/09/2011
Internet address

Keywords

  • Electrooculography
  • Sleep
  • Detectors
  • Electroencephalography
  • Electromyography
  • Medical diagnostic imaging
  • Educational institutions

Fingerprint Dive into the research topics of 'Automatic REM Sleep Detection Associated with Idiopathic REM Sleep Behavior Disorder'. Together they form a unique fingerprint.

Cite this