Comparison of speed and accuracy of ractal methods of determined chaos applied to recognition of sleep phases

Abstract

The article presents the results of use of various fractal methods and their author modifications applied to the automated recognition of sleep phases by computer electroencephalograms. The results of the application of the Grassberger-Procaccia method, Hurst rescaled range method, the author's modification of the method of false nearest neighbors and. the method of approximating entropy for recognition of sleep phases at polysomnography are presented in the work and. compared by relative rate and. accuracy. Brief description of peculiarities of used fractal methods is given in the article. It was showed that use of fractal methods of determinate chaos in synergic approach allows to realize differentiation of sleep stages with relatively high accuracy at significantly reduced time limits with use of only one canal of encephalogram. Reliability of obtained results was estimated. by comparison with the results obtained by certified neurophysiological specialists

Keywords

• somnology
• polysomnography
• electroencephalography
• fractals
• chaos
• synergetics
• hypnogram