Neural representation of the geometry and functionality in a scene

 描述(由申请人提供)：拟议研究的目标是调查大脑如何代表场景几何形状和功能。识别视觉环境是我们与世界日常互动的核心。当我们走进一个新的空间时，我们会迅速认识到是否有一条可循的道路，是否有可跨越的边界，是否有阻碍我们视野和潜在导航的障碍。这一建议的理论框架基于证据，即大脑中一组场景选择区域存在不同但互补的场景表征水平(Park等人，2011；Park等人，2014；Park&Chun，2009；Park，Chun，&Johnson，2010；Park，Intraub，Yi，Widders，&Chun，2007)。PI建议场景几何形状(例如，空间布局、三维场景边界)和功能(例如，可导航性、边界限制)是在这些区域中表示的两个基本场景属性。具体目标：目标1调查大脑是否对垂直边界的存在表现出强烈的敏感性，以及这种敏感性是如何被边界对观察者潜在导航带来的功能障碍所调节的。目的2研究场景可导航性的神经表示，以及这种表示与场景几何表示有何不同。目的3研究真实世界场景的神经表征是否受获取的场景时空背景知识的调制，这些知识对场景的功能很重要。在她的整个目标中，PI测试了人类成年人的内侧颞叶区域，这些区域处理场景和空间信息：特别关注前部和后部的海马旁回和脾后皮质。方法包括单变量和多体素fMRI模式分析(线性支持向量机分类和代表性相似性分析)，以及基于感兴趣区域(ROI)和基于全脑(搜索光)的方法。整个提议的假设和初步结果表明，场景几何表示在海马旁回，而场景功能表示在脾后皮质。

 DESCRIPTION (provided by applicant): The goal of the proposed research is to investigate how the brain represents scene geometry and functionality. Recognizing the visual environment is central to our daily interactions with the world. When we walk into a new space, we rapidly recognize whether there is a path to follow, whether there are crossable boundaries, and whether there are obstacles that block our view and potential navigation. The theoretical framework of this proposal is based on evidence that there are distinct but complementary levels of scene representation across a group of scene-selective regions in the brain (Park et al., 2011; Park et al., 2014; Park & Chun, 2009; Park, Chun, & Johnson, 2010; Park, Intraub, Yi, Widders, & Chun, 2007). The PI proposes that scene geometry (e.g., spatial layout, three-dimensional scene boundary) and functionality (e.g., navigability, limitations of a boundary) are two fundamental scene properties represented in these regions. Specific Aims: Aim 1 investigates whether the brain displays acute sensitivity to the presence of vertical boundaries, and how such sensitivity is modulated by the functional impediment that a boundary presents to the viewer's potential navigation. Aim 2 investigates the neural representation of the scene navigability, and how this representation differs from representation of scene geometry. Aim 3 investigates whether the neural representation of real world scenes is modulated by acquired knowledge about the spatio-temporal context of a scene, which are important for functionality of a scene. Throughout her aims, the PI tests medial temporal lobe regions in human adults that process scene and spatial information: with particular focus on anterior and posterior parahippocampal gyri and retrosplenial cortex. Methods include univariate and multi-voxel fMRI pattern analyses (linear support vector machine classification and representational similarity analysis) in combination with both region-of-interest (ROI) based and whole-brain based (search light) approaches. Hypothesis and preliminary results throughout the proposal suggest that scene geometry is represented in the parahippocampal gyrus, while scene functionality is represented in the retrosplenial cortex.