High temporal precision sometimes emerges from noisy systems. This is often the case in recurrent networks of dynamic elements (from ionic channels to genes or lasers) that can exhibit emergent collective oscillations of substantial regularity even when the individual elements are considerably noisy. This phenomenon is called stochastic coherence and it can be observed in a large number of systems, for example, in dynamics of global climate. However, how noise-induced dynamics at the local level coexists with regular oscillations at the global level is still unclear. Here we studied the slow oscillation regime exhibited by the cerebral cortex network. Slow oscillations constitute a cortical state consisting of periods of activity or neuronal firing that are interspersed with periods of neuronal silence that alternate at a frequency of around 1 Hz (see Figure). Slow oscillations emerge from the recurrent interaction between cortical neurons, making them a network phenomenon.
In a model of the cortical network, we observed that a combination of stochastic recurrence-based initiation with deterministic refractoriness lead to maximum collective coherence for an intermediate noise level, such that noiseinduced dynamics at the local level coexisted with regular oscillations at the global level. Computational analysis of a biologically realistic network model revealed that an intermediate level of background noise lead to quasi-regular dynamics. We verified this prediction experimentally in cortical slices subject to varying amounts of extracellular potassium, which modulates neuronal excitability and thus synaptic noise (see Figure insets). For intermediate noise (or extracellular potassium) levels, the regularity of the slow oscillation was maximum. Furthermore, there was not only a maximum temporal but also spatial regularity. Taken together, these results allow us to construe the high regularity observed experimentally in the brain as an instance of collective stochastic coherence.