Total views : 373

Detection of REM in Sleep EOG Signals


  • Department of Electrical and Electronics Engineering, Selcuk University, Konya, Turkey
  • Department of Computer Engineering, Selcuk University, Konya, Turkey
  • Department of Electrical and Electronics Engineering, Selcuk University, Konya
  • Sleep Laboratory, Faculty of Medicine, Necmettin Erbakan University, Konya, Turkey


Background/Objectives: Sleep staging is very important phase for diagnosing respiration and sleep diseases. Nowadays, Electroencephalogram (EEG), Electromyogram (EMG), Electrooculogram (EOG) signals are particularly used together in studies on sleep staging. Methods/Statistical Analysis: Associating only EOG signals to sleep staging was distinctly purposed. So, this paper deals with extraction features and classifying for determining REM-NREM states from the EOG signals. In this study, left eye (LEOG) and the right eye (REOG) signals were used. After EOG signals were obtained, 21 different features were extracted from LEOG and REOG in time and frequency domain according to rules of American Academy of Sleep Medicine (AASM). Findings: Artificial Neural Networks (ANN) was adopted on features as method of classification with 3-fold cross validation technique and reached conclusion with the maximum test classification accuracy as 88.05%. To obtain higher classification accuracies, Sequential Backward Selection (SBS) method was used. According to results of SBS, number of the best features combination was determined as 13 and the maximum classification accuracy was obtained as 89.62%. The optimum value of hidden layer node number of ANN was determined as 15 for the best features. Application/Improvements: When looking from the viewpoint of percentage of classification accuracy of this study, a result can be seen that is non-negligible value for literature.


Artificial Neural Networks, EOG, Feature Selection, SBS, Sleep Stage.

Full Text:

 |  (PDF views: 301)


  • Uçar MK, Polat K, Bozkurt MR, Bilgin C. Effect of time and frequency domain features in sleep EEG and EOG signals classification. Proceedings of TIPTEKNO‘14, KapadokyaTurkey; 2014. p. 163–66.
  • Redline S. Sleep related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. Sleep. 1999; 22:667–89.
  • Güneş S, Polat K, Dursun M, Yosunkaya Ş. Examining the relevance with sleep stages of time domain features of EEG, EOG, and Chin EMG signals. Proceedings of 14th National Biomedical Engineering Meeting (BIYOMUT), İzmir; Turkey; 2009. p. 1–4.
  • Aungsakul S, Phinyomark A, Phukpattaranont P, Limsakul C. Evaluating feature extraction methods of Electrooculography (EOG) signal for human-computer interface. Procedia Engineering. 2012; 32:246–52.
  • Malaekah E, Patti CR, Cvetkovi D. Automatic sleep-wake detection using Electrooculogram signal. Proceedings of Biomedical Engineering and Sciences (IECBES), Kuala Lumpur; 2014. p. 724–28.
  • Aydogan O, Oter A, Kıymık MK, Deniz T. Automatic classification of sleep stages with artificial neural networks according to visual scoring rules. Proceedings of 23th Signal Processing and Communications Applications Conference (SIU), Malatya: Turkey; 2015. p. 303–6.
  • Liang SF, Kuo CH, Hu YH, Cheng YS. A rule based automatic sleep staging method. Journal of Neuroscience Methods. 2012; 250:169–76.
  • Kohavi R. A study of cross-validation and bootstrap for accuracy estimation and model selection. Proceedings of the 14th international joint conference on Artificial intelligence, San Mateo;1995. p. 1137–43.


  • There are currently no refbacks.

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.