Performance evaluation of an Artificial Neural Network automatic spindle detection system

Sleep spindles are transient waveforms found in the electroencephalogram (EEG) of non-rapid eye movement (NREM) sleep. Sleep spindles are used for the classification of sleep stages and have been studied in the context of various psychiatric and neurological disorders, such as Alzheimer's disease (AD) and the so-called Mild Cognitive Impairment (MCI), which is considered to be a transitional stage between normal aging and dementia. The visual processing of wholenight sleep EEG recordings is tedious. Therefore, various techniques have been proposed for automatically detecting sleep spindles. In the present work an automatic sleep spindle detection system, that has been previously proposed, using a Multi-Layer Perceptron (MLP) Artificial Neural Network (ANN), is evaluated in detecting spindles of both healthy controls, as well as MCI and AD patients. An investigation is carried also concerning the visual detection process, taking into consideration the feedback information provided by the automatic detection system. Results indicate that the sensitivity of the detector was 81.4%, 62.2%, and 83.3% and the false positive rate was 34%, 11.5%, and 33.3%, for the control, MCI, and AD groups, respectively. The visual detection process had a sensitivity rate ranging from 46.5% to 60% and a false positive rate ranging from 4.8% to 19.2%.

[1]  W R Jankel,et al.  Sleep Spindles , 1985, Journal of clinical neurophysiology : official publication of the American Electroencephalographic Society.

[2]  Jean-François Gagnon,et al.  Sleep and quantitative EEG in neurodegenerative disorders. , 2004, Journal of psychosomatic research.

[3]  Nikolaos K. Uzunoglu,et al.  Sleep spindle detection using artificial neural networks trained with filtered time-domain EEG: A feasibility study , 2005, Comput. Methods Programs Biomed..

[4]  G. Buzsáki Rhythms of the brain , 2006 .

[5]  A. Rechtschaffen,et al.  A manual of standardized terminology, technique and scoring system for sleep stages of human subjects , 1968 .

[6]  János Körmendi,et al.  The individual adjustment method of sleep spindle analysis: Methodological improvements and roots in the fingerprint paradigm , 2009, Journal of Neuroscience Methods.

[7]  B Saletu,et al.  Automatic Sleep-Spindle Detection Procedure: Aspects of Reliability and Validity , 1994, Clinical EEG.

[8]  Kemal Polat,et al.  Sleep spindles recognition system based on time and frequency domain features , 2011, Expert Syst. Appl..

[9]  Osman Erogul,et al.  Efficient sleep spindle detection algorithm with decision tree , 2009, Expert Syst. Appl..

[10]  Claudio A. Perez,et al.  Automated Sleep-Spindle Detection in Healthy Children Polysomnograms , 2010, IEEE Transactions on Biomedical Engineering.

[11]  Günther J. L. Gerhardt,et al.  Benchmarking matching pursuit to find sleep spindles , 2006, Journal of Neuroscience Methods.

[12]  C. Smith,et al.  Sleep spindles and learning potential. , 2007, Behavioral neuroscience.

[13]  Nurettin Acir,et al.  Automatic recognition of sleep spindles in EEG by using artificial neural networks , 2004, Expert Syst. Appl..

[14]  M Steriade,et al.  Coalescence of sleep rhythms and their chronology in corticothalamic networks. , 1998, Sleep research online : SRO.

[15]  M. Ferrara,et al.  Sleep spindles: an overview. , 2003, Sleep medicine reviews.

[16]  M. Zervakis,et al.  Time–frequency analysis methods to quantify the time-varying microstructure of sleep EEG spindles: Possibility for dementia biomarkers? , 2009, Journal of Neuroscience Methods.

[17]  Antti Saastamoinen,et al.  Development and comparison of four sleep spindle detection methods , 2007, Artif. Intell. Medicine.

[18]  K. Blinowska,et al.  High resolution study of sleep spindles , 1999, Clinical Neurophysiology.