Stochastic and/or computational modeling of the brain functioning

Abstract

Epilepsy is a very common disease, occurring in about 1% of the world population. It is frequently defined as a state of recurring seizures that may be caused by unbalanced inhibition and excitation in the brain. Recent developments from NeuroMat are putting forward a stochastic theory to account for the brain criticality hypothesis with and without external input. However, there is still a gap to be filled: how does the brain criticality hypothesis relate to the development of neurological disorders? One expects that disorders physiologically expressed as disturbances of healthy brain oscillations, such as epileptic seizures, could be captured by changes in the macroscopic dynamics of neuronal avalanches. This project will address this issue through stochastic models (both mean-field and spatially extended) for a self-organized neuronal network that takes into account epileptic neuronal populations. The neuronal population model should follow from a stochastic simplification of continuum dynamical models, whereas the self-organization should follow the homeostatic mechanisms already present in the literature. The developed models not only will bring knowledge about the functioning of the epileptic brain and its relationship with self-organized critiality, but could be further extended into a data-driven approach to enhance diagnosis and treatment of epilepsy patients.