Mechanism Study Of Neural Information Rocessing Of Biosonar Sound Localization

The information processing capabilities and features of biological neural system receive more and more attention but remains to be explored. As the fundamental units of neural system, neurons and neural circuits are the focus of the relevant researches. Huge experimental data have been collected from present experimental studies, however, for the neural system, which is the information processing center for biological systems, the mechanics of its information detection, encoding, transmission and processing have not been completely understood yet. Recently, significant progress has been made in computational neuroscience, but most of them are limited in the analysis and discussion on specific phenomena observed in neurons or neural network.Considering that the auditory nervous system has neural circuits with sound localization function in its anatomic construction, this thesis applied Hodgkin-Huxley neuron model to analyze and simulate the neural circuit. Theories of circle map and symbolic dynamics were employed to symbolize the simulated neuron output spike trains. And quantitative relationship between the output spike train in biological sonar sound orientation neural circuits and the input stimulation pulse comprising sound localization were investigated. The main findings of the thesis include:1. The simulation of the input stimulus pulse frequency and output pulse sequence of symbols in single neuron demonstrates that, in the circle map with monotonous increasing neuron parameters, the symbol sequence and the frequency of the input stimulus pulses show the simular monotonous changing trend. Moreover, a least change in frequency of 0.04Hz is distinguished.2. By the minimal model simulation of the ITD detection neural circuits, under various stimulation frequencies the monotonic relationship between the time difference of the input spike from both sides and the output spike sequence was investigated, and a time precision of 10μs was achieved. Afterwards, a degeneration model of ITD detection was proposed, which composes of the structure of the classic Jeffress’ ITD model. The detection neurons of each minimum ITD model in the multilayer structure of the degeneration model achieve the identical functions, which consists with the definition of Edelman degeneration.3. An approach of ILD detecting neural circuits was designed and the synaptic connection between neurons is conducted by chemical coupling. The simulation results reveal that with the change of binaural source sound intensity, which is realized by adjusting the frequency of the input stimulus sequence, the time lag in output spikes varies accordingly.4. By comprehensively understanding the physiological features as well as anatomical structure of ITD and ILD, it is found that ILD detection neural circuits have both ITD and ILD detection function. On the basis of the minimal model, a degeneration neural circuit with dual detection function was built and verified by simulation.In summary, this thesis proposed a novel degeneration detection neural circuit for biosonar localization of neural circuits. It was verified by the simulation that the circuit is able to quantitatively describe the input signal. The distance defined in symbolic spatial can distinguish tiny changes of input signals, which are in terms of the time in neural circuit. These findings are in agreement with the features of neural information processing.