Coding For Visual Stimulus Information By Population Of Neurons In Rats Primary Visual Cortex

Vision is the main way of animal perception of external information. For mammals,more than80%of the outside world information is perceived by visual system. Primate visual cortex(V1area) is an important part of the visual system. The vast majority of visual information is transmitted to higher cortex for further cognition after initial integration by V1area. It has a strong sensitivity to visual stimulus orientation and topology of information. For the past few years,many studies have shown that primary visual cortex on visual stimulus information processing and encoding is largely completed by a population of neurons. Therefore, Researches from the perspective of population on V1neurons towards the stimulation and the topology structure of information encoding mechanism to parse visual system information processing mechanism is of great significance.This thesis uses two effective features of V1neurons population response, Spike mean firing rate and local field potential (LFP) gamma band power, based on Lasso logistic regression models to achieve a simple pattern toward the grating on the type or the type of topology discriminant. Details are as follows:1. Design and conduct experiments. Employ grating with different direction or simple graphics of different topology to stimulate visually on anesthesia LE rats, meanwhile, gather Raw Field Potential(RFP) signals of population neurons from V1area using multi-electrode arrays when stimulating rats.2. Rat V1neurons population response analysis. Obtain Spike and local field potential (LFP) signals from raw field potential (RFP) and preprocess the two signals necessarily. Then,extract Spike average fire rate feature and LFPy band power feature,two kinds of effective features,from Spike and LFP signals. And analyze the characteristics of the two facing and topological structure of visual stimuli response characteristics.3.Research on population response features encoding of visual stimulus information. Take Lasso logistic regression as coding model, apply V1area neuron population response Spike average fire rate feature and LFP gamma frequency band power feature and the joint feature of them to the research respectively on coding of the visual stimulation to the orientation information and topology structure information. And judge the results of the coding by cross validation method. In the end, according to the results of coding, analyze the coding effectiveness of all kinds of features generating from different stimulation information and superiority in population coding.