Recurrent Model of Brain Activity Based on the Markov Chain in the Time Domain

The paper proposes a new recurrent model of brain activity, consisting in the sequential activation of nearby dipoles based on a stochastic Markov chain. The proposed model is based on the hypothesis that brain excitation has the character not of a momentary excitation, statically localized in space, but of a certain temporal sequence of localization of excitations, which is proposed to be modeled by a stochastic Markov chain. The Markov chain in the time domain narrows the space of possible solutions to the inverse problem of magnetoencephalography (localization of brain excitations by the observed magnetic field around the head), which increases the accuracy of its solution. A mathematical description of the model and simulation results are presented.