Percolation transition in active neural networks with adaptive geometry

A mathematical model has been proposed for a neural network whose morphological structure varies dynamically depending on activity. This is the property of the so-called structural plasticity typical of developed neural systems of a brain. It has been shown that the spontaneous generation and propagation of a signal in such networks correspond to a percolation transition and the appearance of the connectivity component covering the entire system. Furthermore, adaptive change in the geometric structure of a network results in the clustering of cells and in the reduction of the effective percolation threshold, which corresponds to experimental neurobiological observations.