Quasi-Critical Brain Dynamics on a Non-Equilibrium Widom Line
Is the brain really operating at a critical point? We study the non-equilibrium properties of a neural network which models the dynamics of the neocortex and argue for optimal quasi-critical dynamics on the Widom line where the correlation length is maximal. We simulate the network and introduce an analytical mean-field approximation, characterize the non-equilibrium phase transition, and present a non-equilibrium phase diagram, which shows that in addition to an ordered and disordered phase, the system exhibits a quasiperiodic phase corresponding to synchronous activity in simulations which may be related to the pathological synchronization associated with epilepsy.