Pattern transition of neuronal networks induced by chemical autapses with random distribution

The difference of congenital inheritance and acquired development makes the autaptic distribution in different brain regions variant. To investigate the physiological regulation of autaptic structures on the nervous system, the effects of chemical autapses with random distribution on the dynamics of Newman-Watts small-world neuronal networks are systematically analyzed with the help of three network metrics. The autaptic occurring probability is first introduced to characterize the random distribution of autaptic structures. Numerical results show that the random distribution of chemical autapses can markedly modulate the electrophysiological activities of neuronal networks owing to the self-feedback function of excitatory autapses, not only promoting the transmission of neural signals, but also inducing the network-level stochastic resonance and the transition of network dynamics. Particularly, the autaptic random distribution can make subthreshold or chaotic neuronal networks generate the phase synchronization phenomena and eventually evolve into the synchronous periodic discharge state, which provides a strategy to achieve the complete synchronization through pattern transition. This study reveals the ability of the stochastic characteristics of autaptic structures to alter the evolution of network spatiotemporal patterns, which could contribute to the application of autaptic structures in physiological experiments or artificial neural networks.