Action potentials generated by neurons in the cerebral cortex eventually give rise to conscious sensations. Understanding this process requires both a description of what information is represented in the activity of single neurons, and a description of the mechanism by which that representation is generated. Binocular stereopsis (the ability to perceive depth by combining images from the two eyes) is an attractive model system to study both the nature and mechanism of the cortical representation for several reasons: 1. It seems likely that we can explain the mechanisms that underlie disparity selectivity in single neurons of the primary visual cortex. 2. The psychophysical properties have been extensively studied in humans and monkeys. Many of these properties are not straightforwardly reflected in the activity of single neurons, at least in V1. 3. Extensive computational work offers mechanisms that can bridge the gap between 1) and 2). Binocular stereopsis is possible because the horizontal separation of the eyes means that objects nearer or farther than the fixation point project to different locations on the two retinae. These locations differ principally in their horizontal co-ordinates (horizontal binocular disparity). Recent work from my own laboratory and others has generated a very successful model the underlying mechanism for disparity selective responses. Two important tests of this model were completed this year. First, the model makes a very clear prediction concerning neuronal response rates to stimuli presented monocularly and to binocular stimuli which are uncorrelated between the eyes. Second, according to the model, the shape of the monocular receptive field determines the shape of the disparity-selective response. Both of these experimental projects revealed significant inaccuracies in the current model of disparity selective neurons. Interestingly, a single modification to the model resolves both of these difficulties, and also deals with the only other well-documented failure of the model (weak responses to binocular anticorrelation). Since our modified model is now able to explain all experimental observations on these neurons, it suggests that the model correctly describes at least some aspects of the underlying mechanism. Although the model does an excellent job of accounting for the mechanism of disparity selectivity, the simplicity of the model suggests that disparity tuning might arise by chance in the visual cortex. In fact there is no clear evidence that such neurons are specialized to process naturally occurring disparities. In a search for such evidence, we explored the responses of neurons to disparity applied in different directions. In the past, disparity selective neurons in the cortex have been studied with disparities applied in only one direction (often horizontal), which cannot determine whether the encoding is specialized for processing disparities along the horizontal axis. It is therefore unclear whether or not disparity selectivity represents specialization for naturally occurring disparities. In this project, disparity-selective neurons from the primary visual cortex (V1) of awake fixating monkeys were studied, using isotropic visual stimuli (random dot stereograms). Many combinations of vertical and horizontal disparity were used, characterizing the surface of responses as a function of two-dimensional disparity. Despite the isotropic stimulus, the response surface usually showed elongation along the horizontal disparity axis. Thus these neurons modulated their firing rate over a wider range of horizontal disparity than vertical disparity. This is the first clear demonstration that disparity selective cells are specialized for processing horizontal disparity. This is the clearest evidence to date that these neurons are specialized for a role in depth perception.
Showing the most recent 10 out of 24 publications
