On brain fluctuations and the challenges ahead

The brain can be considered as a thermodynamic machine working out of equilibrium, crossed by a stationary (under normal conditions) flux of energy, that is dissipated through thermal fluctuations. The analogy with a fluid state suggests that long space-time correlations have to be present in such a machine to keep it in working conditions, thus making the brain effectively explore dynamical and statistical regimes characterized by scale-free properties. We argue that such a distinctive feature may emerge from an out-of-equilib-rium thermodynamic approach and also from microscopic models (neural networks) with a suitable combination of phenomenological ingredients.