Automatic classification of sleep stages with artificial neural networks according to visual scoring rules

In this study, Apnea / hypopnea index of less than 15 Obstructive Sleep Apnea patients of sleep stages were scored automatically. For automatic sleep scoring visual scoring system is used EEG, EOG and EMG signals using feedforward neural networks with automatic scoring has been performed. The period...

Full description

Saved in:
Bibliographic Details
Published in2015 23nd Signal Processing and Communications Applications Conference (SIU) pp. 399 - 402
Main Authors Aydogan, Osman, Oter, Ali, Kiymik, Mahmut Kemal, Tuncel, Deniz
Format Conference Proceeding
LanguageEnglish
Turkish
Published IEEE 01.05.2015
Subjects
Artifical Neural Networks
Biological system modeling
Brain modeling
Electroencephalography
Electromyography
Electrooculography
Signal processing
Sleep Scoring
Sleep Stages
Online AccessGet full text

Cover

More Information
Summary:In this study, Apnea / hypopnea index of less than 15 Obstructive Sleep Apnea patients of sleep stages were scored automatically. For automatic sleep scoring visual scoring system is used EEG, EOG and EMG signals using feedforward neural networks with automatic scoring has been performed. The period about 8 hours of time which patients spent in bed sleep has been divided into 30 seconds epochs. According to the 2014 produced by the American Academy of Sleep Medicine criteria to scoring, power characteristics of waves has been derived by using 6 EEG signal, taking from central, frontal and occipital region, 2 EOG signal taking from the right and left eyes and 1 EMG signals taking from the chin. Automatic sleep scoring done by using the 9 signals, gives better results than scoring a single channel. It has been thought that this automatic sleep scoring study done using visual scoring rules prevent loss of time and contribution to sleep scores of the physicians.
ISSN:2165-0608
2693-3616
DOI:10.1109/SIU.2015.7129843