Neural Mechanisms of Visual Crowding

项目摘要 视觉感知是一个主动的信息选择过程。此选择的两个关键组件 过程是空间注意和眼跳运动的机制。注意力使我们能够部署 对显著刺激进行特权处理的神经资源有限；眼跳允许我们带来 兴趣集中在凝视中心进行详细检查。注意力和扫视的动态相互作用是 在不适当地整合上下文信息的外围视觉中尤其突出 被认为是视觉拥挤的潜在机制之一。视觉拥挤是指无法辨认 杂波中的刺激是边缘视觉中目标识别的根本瓶颈。它特别是 对黄斑变性(MD)等中心视力丧失的患者有害，因为这些 个体依赖外周视网膜轨迹(PRL)来完成物体识别任务。拥挤地带，面积在 干扰物影响目标识别的视觉空间沿径向轴明显延长 在正常视力下与中心凹对齐。MD患者拥挤区的系统性重组 沿径向PRL轴的细长构件。尽管进行了数十年的研究，但脑出血的神经基础 拥挤和细长形状的拥挤区域的情况仍不清楚。在这项提案中，我们将架起 这些知识缺口在六个神经亚群的活动背景下--两个细胞类别(抑制性 和兴奋性神经元)在三个层次(浅层、输入层、深层)-这是 感觉板层皮质环路。我们的广泛假设是，亚群体特定活动动态形成了 拥挤背后的语境相互作用以及拥挤的形成和重组的基础 区域。我们建议回答关于这一假设的几个关键问题。什么是层流？ 拥挤区域的语境调节机制(目标1)？周遭的时空动力学是什么？ 层皮质环路中的跳动活动(目标2)？拥挤区的神经机制是什么 在中央视区之后的重组(目标3)？我们将使用高密度层流来实现这些目标 清醒的非人灵长类动物视觉区域V4的神经记录。V4，目标识别中的关键节点 路径，也接收强烈的注意相关和动眼信号，是理想的适合于这些 调查。这些研究的结果将揭示拥挤的神经机制，提供 外围视觉表格处理的广泛框架，并为未来可能的干预研究提供信息 中心视力丧失。

Project Summary Visual perception is an active process of information selection. Two critical components of this selection process are the mechanisms of spatial attention and saccadic eye-movements. Attention allows us to deploy limited neural resources for privileged processing of salient stimuli; saccades allow us to bring stimuli of interest to the center of gaze for detailed inspection. The dynamic interplay of attention and saccades is particularly prominent in peripheral vision where inappropriate integration of contextual information has been proposed as one of the underlying mechanisms of visual crowding. Visual crowding is the inability to identify stimuli amidst clutter and is a fundamental bottleneck to object recognition in peripheral vision. It is particularly detrimental in the case of patients with central vision loss such as macular degeneration (MD), since these individuals rely on a peripheral retinal locus (PRL) for object recognition tasks. Crowding zones, the area in visual space over which distractors affect target identification, are markedly elongated along the radial axis aligned with the fovea in normal vision. Crowding zones undergo systematic reorganization in MD patients with an elongated component along the radial PRL axis. Despite decades of research, the neural underpinnings of crowding and that of the elongated shape of crowding zones remain unknown. In this proposal we will bridge these knowledge gaps in the context of the activity of six neural sub-populations – two cell classes (inhibitory and excitatory neurons) in three layers (superficial, input, deep) – that are fundamental components of the sensory laminar cortical circuit. Our broad hypothesis is that sub-population specific activity dynamics form the substrate for contextual interactions underlying crowding and for the formation and re-organization of crowding zones. We propose to answer several key questions regarding this hypothesis. What are the laminar mechanisms of contextual modulation in crowding zones (Aim 1)? What the spatio-temporal dynamics of peri- saccadic activity in the laminar cortical circuit (Aim 2)? What are the neural mechanisms of crowding zone reorganization following central vision zone (Aim 3)? We will achieve these aims using high-density laminar neural recordings in visual area V4 of awake non-human primates. V4, a critical node in the object recognition pathway, that also receives strong attention related and oculomotor signals, is ideally suited for these investigations. The results of these investigations will uncover the neural mechanisms of crowding, provide a broad framework for visual form processing in the periphery and inform possible future intervention studies in central vision loss.