Self-organized criticality applied in computational neuroscience: relating dynamical states to functional connectivity measures and evaluating critical reliability in in-vitro and in-silico cultures of neural networks

The theory of Self-organized criticality (SOC) applied in neural networks has been widely studied over the past fifteen years and its assumptions have been confirmed in both experimental and theoretical studies. The first work in this area, conducted by Beggs and Plenz, has proved that electrical activity of a population of interconnected neurons can be seen, in regular neurophysiological conditions, as a critical branching process with a branching parameter close to 1 and whose observables follow power law distributions having exponents around -3/2; divergences from such conditions are hallmarks of subcritical or supercritical working states. These and many more results have sustained the hypothesis that the brain operates in a self-organized critical state, that is, there may exist some still unclear phenomena that drive a healthy neural network to work at criticality. In addition to SOC theory, another one of the various fields in which modern Neuroscience has developed so far concerns Functional Connectivity (FC), which is defined as a statistical dependency between neural activities happening in distant points of a network and the idea that different patterns of functional connectivity are linked with the accomplishment of different functions is strongly supported by uncountable studies. Here we try to relate functional connectivity patterns to self-organized critical states in order to see whether the determinants of different dynamical states can be pointed out on the hallmark of functionality in order to better understand the unknown phenomena behind them.