Effects of multiple factors on the transmission of excitatory postsynaptic current signals

The propagation of subthreshold signals is an important part of information exchange in neural systems, but the propagation mechanism of subthreshold signals in neural networks and the influence of internal and external factors on propagation are still unclear. Based on this, firstly, a mathematical model for the transmission of subthreshold excitatory postsynaptic current (EPSC) signals in a multi-layer feedforward neural network is constructed. Each element is a HH neuron model. To simulate the connections between neurons in different layers of the cerebral cortex more closely, a multifactor WS small-world network (MF-WS) is proposed, which can form inter-layer differences by adjusting multiple factors. Then, according to the influence factors of MF-WS small-world network formation, explore its influence on EPSC signal transmission. Through research, the increase of synaptic coupling strength W, the number of neighbors K and the proportion of excitatory neurons R are conducive to the propagation of EPSC signals. With the increase of the respective weights of multiple factors, the optimal noise intensity conducive to the propagation of EPSC signals decreases. Finally, a multifactor-consistent BP neural network (MF-C-BP) is proposed to optimize the synaptic coupling strength and connection mode between layers, which increases the consistency of the neural network pulse discharge sequence after the EPSC signal transmission to real correlation (± 0.30 - ± 0.50) or significant correlation (± 0.50 - ± 0.80). By studying the propagation mechanism of EPSC signal and the consistency of pulse discharge, a more potential mechanism for EPSC signal coding in neural network is provided.