Burst Spike Trains Recognition Based on the Dynamic Properties of Hodgkin-Huxley Neurons and Synapses

The functions of neural system, such as learning, recognition and memory, are the emergences from the elementary dynamic mechanisms. To discuss how the dynamic mechanisms in the neurons and synapses work in the function of recognition, a dynamic neural circuit is designed. In the neural circuit, the information is expressed as the inter-spike intervals of the spike trains. The neural circuit with 5 neurons can recognize the inter-spike intervals in 5-15 ms. A group of the neural circuits with 6 neurons recognize a spike train composed of three spikes. The dynamic neural mechanisms in the recognition processes are analyzed. The dynamic properties of the Hodgkin-Huxley neurons are the mechanism of the spike trains decomposition. Based on the dynamic synaptic transmission mechanisms, the synaptic delay times are diverse, which is the key mechanism in the inter-spike intervals recognition. The neural circuits in the group connect variously that every neuron can join in different circuits to recognize different inputs, which increases the information capacity of the neural circuit group. A neural circuit is designed. The input burst spike trains are decomposed by the selectively responding properties of neurons and the intra-burst periods are learned by the neurons. Then the synaptic delay times are modulated by STDP learning rule, and the inter-spike periods are learned and saved in the delay times. After learning, the specifical neural connection structures are formed to the certain input spike trains and the recognition is finished.