Estimation of partial directed coherence and direction of information flow with auditory evoked potentials

In this paper, existence of information flow between different locations of the brain have been analyzed by multivariate auto-regressive modelling where the auditory evoked potentials were measured from several electrodes. The estimation of the information flow was performed by evaluating the partial directed coherence (PDC) and the directed transfer function (DTF) which are function of frequency within the values 0 and 1. When the post-stimulus time interval was analyzed it was observed that both of the methods illustrates the frequency localization where the information flow occurs and the estimation could be performed by given methods. The frequency resolution of the methods depends on the sampling rate and model order. It was determined that the given methods are matched with each other.