DISCRIMINATION OF SLEEP STAGES - A COMPARISON BETWEEN SPECTRAL AND NONLINEAR EEG MEASURES

During recent years methods from nonlinear dynamics were introduced in to the analysis of EEG signals. Although from a theoretical point of v iew nonlinear measures quantify properties being independent from conv entional spectral measures, it is a crucial question whether in practi ce nonlinear EEG measures yield additional information, which is not r edundant to the information gained by spectral analysis. Therefore, we compared the ability of several spectral and nonlinear measures to di scriminate different sleep stages. We evaluated spectral measures (rel ative delta power, spectral edge, spectral entropy and first spectral moment), and nonlinear measures (correlation dimension D2, largest Lya punov exponent L1, and approximated Kolmogorof entropy K2), and additi onally the stochastic time domain based measure entropy of amplitudes. For 12 healthy subjects these measures were calculated from sleep EEG segments of 2:44 min duration, each segment unambiguously correspondi ng to one of the sleep stages I, II, SWS and REM. Results were statist ically evaluated by multivariate and univariate analyses of variance a nd by discriminant analyses. Generally, nonlinear measures (D2 and L1) performed better in discriminating sleep stages I and II, whereas spe ctral measures showed advantages in discriminating stage II and SWS. C ombinations of spectral and nonlinear measures yielded a better overal l discrimination of sleep stages than spectral measures alone. The bes t overall discrimination was reached even without inclusion of any of the spectral measures. It can be concluded that nonlinear measures yie ld additional information, which improves the ability to discriminate sleep stages and which may in general improve the ability to distingui sh different psychophysiological states. This confirms the importance and practical reliability of the application of nonlinear methods to E EG analysis.