Correlates of perceived size in V1 neurons

点击翻译按钮获取中文摘要

The goal of this project is to use behavioral and single-unit recording techniques to understand the role of three-dimensional (3D) context in the neural processing of object size. Estimating the size of an object highlights a fundamental computational problem in vision: the inherent ambiguity in the retinal image. For any given retinal image there are an infinite number of combinations of object sizes and viewing distances that could give rise to that image. A fundamental challenge is to understand how distance information - which is not explicitly represented in the retinal image - is combined with retinal size information to achieve an accurate and stable representation of object size. The overarching framework of this proposal is that distance information is combined with retinal size information in early stages of the visual system. Convergent evidence from behavioral, fMRI, and ERP experiments in humans have shown that object size is represented in the primary visual cortex. These results are inconsistent with the primary visual cortex passively reflecting the stimulus on the retina, but rather suggest that high-level signals related to 3D scene interpretation may be fed back to primary visual cortex to adjust the amount of tissue allocated to represent visual objects. To extend these findings, our proposed experiments pursue these results in an animal model using psychophysical measurements and single-neuron electrophysiological recordings. Two specific aims are proposed: 1) single-unit recording to investigate the changes in receptive field structure of individual visual cortical neurons as a function of 3D context and 2) psychophysical measurements of size illusions in the same animal model used in specific aim 1. Together these specific aims will shed light on the processes that underlie size perception, which are critical for interacting with a 3D world, and will provide a bridge between vision studies in human and non-human subjects.