Electrical Model Of The Neural Information Conduction

Nervous system is a complex network composed by a large number of neurons which conduct the neural information mainly through the membrane voltage and synaptic current. Constrained by the performance of general-purpose computer, software simulation can not satisfy the biological neural network research requirements. Instead of reading the program, digital ASIC devices can complete the computation in parallel mathematics. Implementing the electrical model of nervous system with digital ASIC can improve the computational efficiency significantly. It is an effective method to study the electrophysiological properties of the nervous system. Based on field programmable gate array (FPGA) technology, the electrical model of neural information conduction was built in this paper, and the experimental results show the validity of the model.Based on the pipeline idea, the data path in the FPGA can solve multiple ordinary differential equations simultaneously. Firstly, in this paper, the strict mapping between the pipeline data path and the ordinary differential equation was established by the method of pipeline operator, and the Multi-Process Pipeline Model of ordinary differential equation was built. Secondly, the necessary condition to maintain the neural model and biological neurons in the same time scale was deduced by analyzing the time and space distribution relationship of the state variables in the Multi-Process Pipeline Model. Finally, the solution to implement the coupling ordinary differential equations with pipeline data path was given.In this paper, an experimental platform based on FPGA technology was designed. It was composed of FPGA computing core; USB data upload system and the AD, DA conversion circuit. The delay of the AD and DA conversion process will impact the real-time nature of the platform. Two sequential control methods were proposed to ensure that the AD and DA processes can be completed in one iterative step. According to the electrophysiological properties of Morris-Lecar neurons and the neural network, the electrical model of Morris-Lecar neurons and the network was implemented with the Multi-Process Pipeline Model. The firing patterns and synchronization phenomena were analyzed through the electrical model, indcating the correctness and feasibility of the method.Hippocampus is an important brain tissue to cognitive, emotional and other important brain functions. The electrical model of hippocampus neurons in CA3 area was built by the Multi-Process Pipeline Model mehtod in this paper. The electrophysiological properties of the hippocampus neuron were analyzed by the electrical and the results show that the synchronization phenomenon exists in the hippocampus neural network.An effective means to research the production and conduction process of neural information was proposed in this paper. The dynamics of biological nervous system can be implemented quickly and accurately with the electrical model. It supplies a recoverable experimental subject for the research of acupuncture information conduction.