The Perceptual Learning In Sinusoidal Gratings Orientation Discrimination And Contrast Detection In Cat

The performance of visual perception can be improved by repeat training. Through analyzing the model of animal perceptual learning, and with the method of electrophysiology, cell and molecular biology, morphology and functional imaging techniques, we can probe the mechanism of learning and memory. In this paper, we introduced the method and set of cat perceptual learning and research the character of orientation discrimination and contrast detection of cat, and we also research the effect of visual perceptual learning on the response of the Dorsal Lateral Geniculate Nucleus(dLGN) neuron with the method of single-neuron in vivo recording techniques, by which to the ends of knowing the mechanism of neuron in the perceptual learning. Perceptual learning in orientation discrimination and contrast detection was first observed by psychophysics method in cat in the following two experiments. The effect of perceptual learning is obvious: the performance is greatly improved after repeat training. However there are different end we observed between the two experiments. In the contrast sensitivity discrimination task, monocular learning could significantly improve contrast sensitivity for both eyes, which was spatial frequency dependence. But there was a significant inter-eye difference, which is that the contrast of sensation of trained eye is proved more than untrained eye, indicating an incomplete but evident inter-eye transfer of the learning. In the orientation perceptual learning experiment, there are no obvious inter eye transfer between the training eye and the untraining one while in detection of orientations with consecutively changing difference task, but it is reverse in the discrimination on fixed orientation. The partial inter-eye transfer of contrast detection task and SF-tuned specificity of learning in cats are similar to that observed in human subjects. Another interesting phenomenon was found in both experiments: both of the learning process can be segmented into a fast learning process at first and a slow and long process at last, which was similar to the report of human perceptual learning.The first process probably is the end that the generation of learning effects on the perceptual task, and the latter process probably is a process of the neural plasticity. Using single-neuron in vivo recording techniques, we research the effect of orientation learning on dLGN neurons orientation specificity properties. Extracellularly recording the responses of dLGN neurons to sinusoidal gratings with different orientations showed that optimum orientations of dLGN neurons in trained cats, compared with that in untrained cats, did not shift in favour of the trained orientation. For cells with receptive fields located within 15°visual acuity away from the foveae, there are no difference in orientation bias and firing rate at trained orientation between the trained cats and untrained cats. All above indicate that oriented grating discrimination learning does not lead to a shift of orientation sensitivity for dLGN neurons in the cat. The specificity of performance improvement at trained orientation may be related to neural plasticity of orientation encoding in the visual cortex.