Processing of spatial information in V1 through fixational eye movements

点击翻译按钮获取中文摘要

PROJECT SUMMARY This project focuses on cortical mechanisms underlying the information-processing roles of fixational eye movements (FEM's) in vision. FEM's, the microscopic “jittering” of the eyes that is constantly present when gaze is not being shifted – have long been regarded as a “bug”, i.e., a source of blurring that must be overcome. But based on work in collaboration with Michele Rucci's lab, it is now clear that this jittering is not a bug, but rather, a feature, i.e., a crucial information-processing step. The specific characteristics of fixational eye movements allow for efficient representation of spatial information in the time domain, allowing later neural processing to use dynamics, such as neural synchrony, to extract visual features. This fundamental shift in viewpoint has many important implications, including: (i), since fixational eye movements are under oculomotor control, they provide a mechanism by which the visual system's properties can be tuned according to task, and (ii), since the normal pattern of FEM's is essential for normal visual sensitivity, disease- and age-related changes in oculomotor function necessarily lead to alterations in visual perception that were previously thought to be purely sensory in origin. Our previous work establishes the importance of fixational eye movements via perceptual experiments in man and modeling studies, but provides no knowledge of the physiological underpinnings. In Aim 1 we will establish a non-human primate system that will enable us obtain this knowledge, and in Aim 2, we will use it to address fundamental questions concerning mechanism. Specifically, in Aim 1 we will integrate several technologies that we and our collaborators are experts in, but which have never been combined, including: implantation of chronic high-density recording arrays in macaque visual cortex, construction of a high- resolution eye tracking system based on scleral coils, construction of a custom visual display system capable of selectively stabilizing portions of a visual image based on these coil signals, online analysis of neural activity for trial-by-trial comparison with performance, and offline reconstruction of receptive fields and the position and movement of the eyes at a higher resolution than possible with scleral coils alone. In Aim 2, we will use this system to record from V1 during visual discrimination tasks, under conditions of full and partial retinal image stabilization. These experiments will determine whether (a) FEM's act to support high acuity vision via generation of luminance transients, and (b) whether efferent-copy and/or motor-planning signals are necessary to make use of this information. Our overall hypothesis is that the luminance transients create distinct temporal patterns of activity in V1 neurons, and that these temporal patterns underlie high-acuity vision. More broadly the proposal that the visual system uses oculomotor behavior to represent space through time implies that vision is an active process, more analogous to other senses (e.g. somatosensation) than commonly postulated, a conceptual shift with major implications for studying visual function and dysfunction.