Representation of attentional priority for visual features in the human brain

人脑视觉特征的注意力优先级表示

• 批准号: 10440619
• 负责人: Taosheng Liu
• 金额: $ 36.83万
• 依托单位: MICHIGAN STATE UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-30 至 2026-06-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10440619
• 关键词: Affect; Alzheimer' s Disease; Anatomy; Architecture; Area; Attention; Attention Deficit Disorder; Attention deficit hyperactivity disorder; Attentional deficit; Auditory; Automobile Driving; Basic Science; Behavior; Behavioral; Biological Models; Brain; Categories; Characteristics; Clinical Research; Code; Complement; Computer Models; Data Analytics; Diagnosis; Dimensions; Disease; Dissociation; Dorsal; Environment; Foundations; Functional Magnetic Resonance Imaging; Goals; Health; Human; Impairment; Influentials; Literature; Location; Measures; Mediating; Methodology; Methods; Mission; Modeling; Nature; Neuropsychology; Neurosciences Research; Perception; Population; Process; Property; Psychophysics; Radiology Specialty; Reading; Research; Research Project Grants; Resolution; Role; Scanning; Selection Criteria; Signal Transduction; Social Interaction; Stimulus; Stimulus Deprivation-Induced Amblyopia; Stroke; Structure; Techniques; Testing; Theoretical Studies; Time; Variant; Visual; Visual Perception; Visual attention; Visual system structure; Work; advanced analytics; autism spectrum disorder; base; cognitive control; cognitive neuroscience; feature selection; information processing; innovation; insight; multimodality; neglect; nervous system disorder; neuroimaging; neuromechanism; neurotransmission; novel; relating to nervous system; response; screening; selective attention; somatosensory; support network; theories; visual information; visual stimulus

PROJECT SUMMARY The environment contains far more information than the brain can process at once. Visual attention helps us cope with such information overload by selectively processing relevant information. In many situations, humans need to select arbitrary features in a scene. Theories of attention have proposed that such selection is mediated by a priority representation that encodes the relative importance of each visual stimulus in the scene. However, much remains unknown regarding how the brain computes and maintains attentional priority for features. Our long-term goal is to understand how the brain selects different types of information via population neural activity to serve goal-directed behavior. In this project, we will examine the neural basis of two basic properties of feature attention: its resolution and capacity. We hypothesize that distinct areas in the dorsal frontoparietal network encode priority information with different resolution and capacity limit, supported by distinct neural population activity profiles. We will test this overall hypothesis by pursuing three specific aims. First, we will establish functional specializations in frontoparietal areas in representing feature priority with different levels of resolution. Second, we will examine the nature of priority signals that gives rise to the capacity limit in attending to multiple stimuli. Third, we will quantify the dimensionality of priority signals and examine how neural dimensionality determines the resolution and capacity of the priority representation. The proposed research is expected to significantly advance our understanding of how the brain selects visual features, in terms of the neural machinery and computational principles that enable such selection. A deeper understanding of how the brain selects visual features will provide important constraints for theories and models of attention and can potentially transform our understanding of visual information processing and cognitive control. The research project is innovative both in terms of conceptual and methodological advances. Conceptually, the project will test novel hypotheses regarding the functional dissociations in frontoparietal cortex and the underlying computational principles of neural coding. Methodologically, the project employs a multi- modal approach including behavioral, neuroimaging, and neuroperturbation techniques, complemented by advanced data analytical and computational modeling methods, to gain fundamental insights into the brain mechanisms of visual attention.

项目摘要 环境包含的信息远远超过大脑一次可以处理的信息。视觉注意 帮助我们通过有选择地处理相关信息来科普这种信息过载。在许多 在某些情况下，人类需要选择场景中的任意特征。注意力理论已经提出 这种选择是由优先级表示来介导的，该优先级表示对以下各项的相对重要性进行编码： 场景中的每一个视觉刺激。然而，关于大脑是如何 计算并维护特征的注意力优先级。我们的长期目标是了解 大脑通过群体神经活动选择不同类型的信息， 行为在这个项目中，我们将研究特征注意力的两个基本属性的神经基础： 其分辨率和容量。我们假设背侧额顶叶网络中的不同区域 以不同的分辨率和容量限制编码优先级信息，由不同的神经网络支持 人口活动概况。我们将通过追求三个具体目标来检验这一总体假设。第一、 我们将在额顶叶区域建立功能特化，以表示特征优先级， 不同的分辨率。其次，我们将研究优先信号的性质， 注意多种刺激的能力限制。第三，我们将量化优先级的维度 信号，并检查神经维度如何确定优先级的分辨率和容量 表示.这项拟议中的研究预计将大大推进我们对如何实现这一目标的理解。 大脑根据神经机制和计算原理选择视觉特征， 这样的选择。更深入地了解大脑如何选择视觉特征将提供 注意力理论和模型的重要限制，并可能改变我们的 理解视觉信息处理和认知控制。该研究项目是 在概念和方法上都有创新。从概念上讲，该项目将 测试关于额顶叶皮层功能分离的新假设， 神经编码的基本计算原理从方法上讲，该项目采用了多种方法， 模式方法包括行为、神经成像和神经扰动技术， 辅以先进的数据分析和计算建模方法， 视觉注意力的大脑机制的基本见解。