Detrended Fluctuation Analysis Features for Automated Sleep Staging of Sleep EEG

Detrended Fluctuation Analysis Features for Automated Sleep Staging of Sleep EEG

Amr F. Farag (Department of Systems and Biomedical Engineering, Cairo University, Giza, Egypt, & Department of Biomedical Engineering, Shorouk Higher Institute of Engineering, EL-Shorouk, Egypt) and Shereen M. El-Metwally (Department of Systems and Biomedical Engineering, Cairo University, Giza, Egypt)
DOI: 10.4018/ijsbbt.2012100104
OnDemand PDF Download:
No Current Special Offers


An accurate sleep staging is crucial for the treatment of sleep disorders. Recently some studies demonstrated that the long range correlations of many physiological signals measured during sleep show some variations during the different sleep stages. In this study, detrended fluctuation analysis (DFA) is used to study the electroencephalogram (EEG) signal autocorrelation during different sleep stages. A classification of these stages is then made by introducing the calculated DFA power law exponents to a K-Nearest Neighbor classifier. The authors’ study reveals that a 2-D feature space composed of the DFA power law exponents of both the filtered THETA and BETA brain waves resulted in a classification accuracy of 93.52%, 93.52%, and 92.59% for the wake, non-rapid eye movement and rapid eye movement stages, respectively. The overall accuracy of the proposed system is 93.21%. The authors conclude that it might be possible to build an automated sleep assessment system based on DFA analysis of the sleep EEG signal.
Article Preview


Sleep is not just a constant state controlled by metabolic needs for the body being at rest. Instead, sleep consists of different well-defined sleep stages, namely, wake (WK), rapid eye movement (REM) and non-REM sleep. In a normal restorative sleep, these stages follow a well-structured temporal order (Carskadon et al., 2000).

For more than 40 years, visual assessment of wakefulness and sleep in clinical sleep studies has been based on standard manual of Rechtschaffen and Kales (R&K) (Rechtschaffen et al., 1968). Although this manual is considered the gold standard inside sleep research community, a considerable amount of research has been carried to define methods that would give a more detailed and accurate sleep description of sleep macrostructure and overcome the known limitations of the R & K manual (Himanen et al., 2000; Hasan et al., 1996; Penzel et al., 2000).

During recent decades, a multitude of methods aiming at objective, continuous-scale quantification of sleep depth have been presented (Hasan et al., 1996; Penzel et al., 1991, Kemp 1993). Most of the important early findings of clinical sleep medicine were based on period analysis, which makes it possible to carry out time-frequency analysis even visually for properly band-pass filtered data (Hasan et al., 1996). Hjorth parameters were introduced to characterize amplitude, time scale and complexity of the EEG through time-domain operations and were exemplified to be applicable in the analysis of objective sleep depth (Hjorth, 1970). More recently, more studies on sleep staging have been conducted including: at least stochastic complexity measures (Rezek et al., 1970), relations of certain spectral bands (Jobert et al., 1994; Dimpfel et al., 1998; Hammer et al., 2001), models on EEG micro-continuity (Mourtazaev et al., 1995), Hidden Markov Models (Flexer et al., 2005), segmentation approaches (Kaplan et al., 2001), k-means clustering based feature weighting combined with K-Nearest Neighbor and decision tree classifier (Gunes et al., 2010), and Fuzzy logic combined with genetic algorithm (Jo et al., 2010).

Complete Article List

Search this Journal:
Open Access Articles: Forthcoming
Volume 3: 1 Issue (2015)
Volume 2: 4 Issues (2013)
Volume 1: 4 Issues (2012)
View Complete Journal Contents Listing