GCSNN: Glia-Coupled Spiking Neural Network

Recent research has uncovered the profound importance of glial cells in the brain. As supporters and regulators of the central nervous system, they play a crucial role in the normal functioning of the neural system through their interactions with neurons. They can conduct bidirectional communication of information with neurons, and have the function of enhancing or inhibiting synaptic signal transmission. Taking inspiration from this biological characteristic, the objective of this paper is to develop a spiking neural network model that incorporates glial cell coupling. The model utilizes a spiking neural network as its foundation and adopts a tripartite synapse model to design an algorithm for updating neuron states. By leveraging glial cells to enhance or suppress stimulation, synaptic transmission is improved. The experiments show that incorporating glial cell regulatory mechanisms into spiking neural networks leads to improvement in the accuracy of test set on both conventional datasets and neural morphological datasets.