Statistical study of the EEG in motor tasks (real and imaginary)

EEG is one of the techniques more used to assess the extent of damage from these deficiencies and even to find solutions such as rehabilitation or limb replacement using a bionic prosthesis, through brain-computer interface. That is why it is of vital importance that we understand the functioning of the primary motor cortex of the brain in the control of real/imaginary tasks. From Physionet database, with two-minute EEG recordings in three different experiments (real/imaginary), we applied DFA and DCCA methods to find auto-correlations and cross-correlations. DFA method was capable of quantitatively describing similarities when the brain performs the same motor task, and show there are three time-scales. After, in order to compare the fluctuation amplitude of an EEG signal in relation to the other channels and measure these cross-correlations, we applied.log FDFA function and.DCCA coefficient. Thus, choosing the F-3 channel (front) as the reference, we identified generally that: Delta log F-DFA[F-3 : xx] >= 0 and rho(DCCA) > 0. The channels: C-z, F-6, T-9, and T-10, are those that have a higher level of DCCA cross-correlation, if compared to the channel F3. The time scale II, with 16 < n <= 723, is the one with rho(DCCA) maximum. Finally, the statistic applied in this paper, based on the DFA-method, proved to be an excellent candidate for studies of motor functions in the brain computer interface area.