Análise tempo-frequência de sinais eletroencefalográficos normais antes, durante e após estímulo musical

Abstract

Surface electroencephalography (EEG) recording has important characteristics in relation to high frequency oscillations, which are induced in cognitive tasks engaged in perception, attention and memory which reflect cortical activation. However, there are still few studies that analyze the behavior of high frequencies before a musical stimulation. Usually the available surveys, analyze only the Clinical Band, that is, the frequency range of 1 to 30 Hz, being excluded from this analysis the Gamma and Super-Gamma rhythms. Therefore, the objective of this work is to analyze, at high frequencies, EEG signal power variation in the presence of musical stimulation, in which the presented stimulus is unknown to the population. In order to make possible the proposed analysis, a quantizer was developed, called Normalized Power Percentage (PPN), which has the same principle of Spectral Density of Power, with the difference that this value is necessarily normalized. In addition, for the work presented here, the analysis was done for signals acquired before the musical stimulation, part eyes closed, during the stimulation, part music and after the end of the stimulation, part post music. To verify if there are differences in the three parts analyzed, a statistical analysis of the data was performed and the PPN variation of the signal in time was also observed from the time-frequency diagrams. From the results obtained it was possible to verify that the high frequencies, Gamma and Super-Gamma, have a better distinction of the different parts, and that in general, especially for the fastest rhythms, the value of PPN increased during musical stimulation. Finally, it was also noted that after the withdrawal of the stimulus there is a gradual decay of this value and that this is independent of the region of the cerebral cortex analyzed.