Research On Extracting Features Of Spike Train In Primary Visual Cortex And Neuronal Orientation Preference

The vision is one of the main paths through which animals obtain information from outside world. Above 80% of the outside world information are obtained by vision. Animal visual systems have become one of the Information Processing System, which has the most complete function and the most outstanding mechanism, after the number hundred million years. The study of information processing system of animals is the major mutual research of psychology, Neuroscience, Intelligence Science and computer science. The area V1 acts as the primary visual cortex almost receives all of the input coming from the retina. In addition, the area V1 contains nearly all the feature of the natural image without redundancy. Therefore, V1 is the best location to collect the visual signal and to make research on how the visual cortex encodes the visual information.This paper took rat as our experiment object and mainly studied the corresponding relationship between the orientation and response characteristics in the primary visual cortex under the stimili of raster. The following is the main content of this paper:First, a detailed experimental scheme is designed, including the establishment of the experiment platform, the design of surgical operation scheme,the development of visual stimulation system and the discussion of the visual stimulation scheme. All of these work would provide neurophysiological support for the research on the extraction of VI response characteristics and the relationship between those features and orientation of grating.Second, the visual signal was collected in rat V1 under the stimulation of raster according to the above experimental scheme, and several pre-analysis were carried out on those raw data, including the detection of the action potential and the sorting of the spike, In addition, a detailed sorting algorithm proposed and used in this paper is given next which is based on wavelet transformation and Fuzzy-C mean cluster. All of those work prepared for extracting the response feature of singal neuron and establishing the relationship between those features and orientation of grating.Third, the response charateristics of V1 under the stimulation is compared with that of V1 before the stimulation in the following three aspects——firing rate,response latency and the synchronization of each two neurons. And it shows that those response characteristics all changed remarkably after the beginning of the stimulation. Reverse correlation is used to map the orientation tuning curve of each neuron. In addition, the data of big sample was analyzed by using some statistical methods such as one way variance analysis and multiple comparisons.Finally, the corresponding relationship between stimulus and response was established based on covariance function. Moreover the orientation of the stimulus was predicted accord correlation. The result shows that this method can predict some of the orientations successfully.