Neural network model to simulate neuronal responses of Aplysia gill-withdrawal reflex

Simultaneous recordings of many neuron's activities were done during gill withdrawal reflex of a sea snail, Aplysia, by multichannel optical recording. During an optical experiment, five electrical stimuli were given to the siphon nerve to activate the siphon sensory neurons. Detected neurons were classified into 5 groups by the pattern of their action potential trains. The neurons in one of the groups showed burst-like high frequency firing after the second or third electric stimulus. These burst-like activities were correlated with gill movement; the burst occurrence was found to accompany its enhancement. Action potential trains of each type of neurons were simulated by a neural network model, whose partial structure was inferred by the information theoretic analysis previously reported. Roles of the neurons in each group in a neural network were discussed.