Disparity processing in human visual cortex

点击翻译按钮获取中文摘要

The fact that we view the world through two eyes provides us with the computational basis for exquisite sensitivity to the structure and layout of the three-dimensional world. This sensitivity is derived, in large part, through a combination of retinal disparity and motion cues. Our understanding of the computations and cortical mechanisms that underlie mature depth perception is still rudimentary and there is a lack of neural developmental data regarding sensitivity to depth cues outside of the earliest visual areas and youngest ages. In the first Aim, dynamic random dot stereograms will be used to measure the complete developmental sequence for a critical performance limit of the stereoscopic visual system – the minimal disparity that can be detected (stereoacuity). These limits will be measured in developing infants, children and adults using Steady-state Visual Evoked Potentials and will be related to measurements of spatial contrast sensitivity in the same participants. The comparison will allow us to determine whether low-level stimulus visibility is the critical limiting factor in the development of stereopsis or whether more central limits are involved. In the second Aim, sensitivity to binocular motion cues will be studied across development, as these cues provide independent information for scene layout and may involve mechanisms different from those responsible for coding binocular disparity. Preliminary data suggest that a binocular motion mechanism sensitive to differences in velocity between the two eyes that is present in the adult is absent in infants. This Aim will provide the complete developmental sequence for this system and will determine its relationship to monocular motion sensitivity. The third Aim will use functional Magnetic Resonance Imaging (fMRI) to localize disparity and binocular-motion-related activity within identified visual areas of the healthy adult visual system. These data will provide evidence regarding opponent binocular mechanisms we hypothesize form the foundation of figure-ground segmentation/depth estimation and that are selectively immature in infancy. The fMRI data will also identify areas most likely to support the perception of depth from disparity and binocular motion cues. The data sets we will provide will be the first functional measurements of the development of sensitivity to disparity, motion and contrast made using a consistent and sensitive methodology throughout the development of key mechanisms of spatial vision. This research addresses fundamental questions about how changes in cortical response organization over human visual development allow the visual system to encode the spatial layout of the environment. The timing of the maturation sequences of these processes is directly relevant to developmental disorders of visual processing – especially strabismus and amblyopia. Importantly, the methods used here to study normal visual development are readily adaptable for use in clinical research. The large-scale EEG, fMRI and structural MRI data sets that will be generated by the project will be documented and deposited for others to use for testing models of visual development and for refining new multi-modal imaging approaches.