Ya.A. Turovsky, S.D. Kurgalin, A.A. Vahtin
This paper presents a method that develops the technology of processing the biomedical signals based on wavelet analysis. The paper demonstrates that this method allows obtaining a physiologically meaningful information indicated in the dynamic of the behavior of the local frequency maxima scalogramms, which are obtained during the processing of electroencephalograms (EEG). There are two types of distribution parameters of local maxima, related, presumably, to the activation, synchronization, inhibition, and desynchronization of the brain neuronal ensembles producing EEG waves. Different features of the drift of the local maxima in different frequency bands are demonstrated.
Thus, the shifts of oscillator peaks’ coordinates of frequency are observed less frequently in the γ-rhythm ranges, compared to the more low-frequency ranges, due to much poorer resolution of the applied wavelet in respect of frequency at higher frequencies. Nevertheless, the presence of frequency drift to both the decreasing and increasing shows that the frequency of peaks of γ-rhythms, which are associated with cortical activity of the brain and cognitive functions, vary considerably stronger than for the α-and β-rhythms. It is noted that the restricted use of this method is associated with low resolution of the examined signals in respect of frequency for the types of applied wavelets, as well as for the real time processing systems for the signals. The latter is associated with a long duration of the analyzing wavelet functions in a temporary space, which requires the accumulation of the signal of a certain length before it becomes possible to start processing it. The obtained results allow us proceeding with further detailed studies of the physiological mechanisms of the discovered phenomena via using the high-density EEG data.