Comparative analysis of EEG by DFA and wavelet analysis for the estimation of DOA

In monitoring depth of anaesthesia for several decades, a number of methods have been elucidated and developed for the assessment of level of hypnosis under the effect of anaesthesia. The application of anaesthetic agents shows significant effects on electroencephalograph (EEG) waveform. In this paper we firstly estimated the depth of anaesthesia using detrended fluctuation analysis and wavelet analysis in order to characterize the patient state. Detrended fluctuation analysis assessed and examined the scaling behaviour of EEG in order to assess scaling information & long range correlations in time series. This scaling behaviour exponent consists in performing a linear regression fit of a scale-dependent quantity versus the scale in a logarithmic representation. This includes the Detrended Fluctuation Analysis (DFA). But in time domain, analysis of EEG is complex and time-frequency approach is needed. Therefore, Wavelet analysis, in particular, provides means of time-frequency localization of the information. Time resolution is improved which allows detection of the time of its occurrence. In Discrete wavelet transform, EEG signals were decomposed into sub-bands. A mathematical Probability density of each sub-band of each EEG segment was calculated according to number of wavelet coefficients in order to obtain uniformly time distributed atoms of energy across all the scales. This second method provides more robust results and can be applied to more general models.