Paper detail

Network algorithmics and the emergence of synchronization in cortical models

When brain signals are recorded in an electroencephalogram or some similar large-scale record of brain activity, oscillatory patterns are typically observed that are thought to reflect the aggregate electrical activity of the underlying neuronal ensemble. Although it now seems that such patterns participate in feedback loops both temporally with the neurons' spikes and spatially with other brain regions, the mechanisms that might explain the existence of such loops have remained essentially unknown. Here we present a theoretical study of these issues on a cortical model we introduced earlier [Nathan A, Barbosa VC (2010) Network algorithmics and the emergence of the cortical synaptic-weight distribution. Phys Rev E 81: 021916]. We start with the definition of two synchronization measures that aim to capture the synchronization possibilities offered by the model regarding both the overall spiking activity of the neurons and the spiking activity that causes the immediate firing of the postsynaptic neurons. We present computational results on our cortical model, on a model that is random in the Erdős-Rényi sense, and on a structurally deterministic model. We have found that the algorithmic component underlying our cortical model ultimately provides, through the two synchronization measures, a strong quantitative basis for the emergence of both types of synchronization in all cases. This, in turn, may explain the rise both of temporal feedback loops in the neurons' combined electrical activity and of spatial feedback loops as brain regions that are spatially separated engage in rhythmic behavior.