Competition Model of Attention and Memory

注意力和记忆力的竞争模型

• 批准号: 7372449
• 负责人: MARY C POTTER
• 金额: $ 27.81万
• 依托单位: MASSACHUSETTS INSTITUTE OF TECHNOLOGY
• 依托单位国家: 美国
• 财政年份: 1992
• 资助国家: 美国
• 起止时间: 1992-08-01 至 2010-07-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7372449
• 关键词: Accident and Emergency department; Accidents; Accounting; Address; Air; Attention; Automobile Driving; Behavioral; Biological Neural Networks; Blindness; Blinking; Cities; Collaborations; Complex; Computer Simulation; Conscious; Crime; Cues; Data; Detection; Disease; Electroencephalography; Emergency Situation; Environment; Event; Eye; Follow-Up Studies; Investigation; Letters; Life; Light; Link; Location; Mediating; Medical; Memory; Modeling; Monitor; Numbers; Pattern; Performance; Process; Rate; Reading; Regulation; Relative (related person); Reporting; Research; Research Personnel; Resources; Series; Short-Term Memory; Signal Transduction; Sorting - Cell Movement; Specific qualifier value; Spottings; Stimulus; Stream; Structure; Symptoms; Testing; Visual; Visual Fields; Visual Pathways; Visual attention; Visual system structure; improved; magnocellular; memory encoding; neuroimaging; parvocellular; research study; response; trafficking

DESCRIPTION (provided by applicant): The overall aim of this proposal is to understand and model the processes that underlie the formation of conscious, reportable memories of otherwise fleeting visual representations. When we are looking for something, we can pick out that target surprisingly quickly when our eyes light on it. This is shown by presenting items one after the other at rates of 10 or even 20 a second using rapid serial visual presentation (RSVP). Such experiments with sequences of letters, words, or pictures have shown that the first target is easy to spot, but then it is strikingly difficult to detect a second target for the next half second (500 ms), an effect called an attentional blink. A complex sequence of processes occurs in that 500 ms, starting with detection of the first target, followed by a brief period of enhanced attention called transient attention which rapidly draws attentional resources to the location of the target. After this initial burst of attention, the blink occurs, making it very difficult to spot a new target anywhere in the visual field for 200-500 ms afterwards. Surprisingly, this blink is entirely eliminated by presenting an unbroken stream of targets. Thus a string of targets keeps attention engaged, but if there is a gap in the targets, attention is switched off. This effect suggests that there is a struggle between top down and bottom up control for attention. To understand these dynamics, a computational model of the temporal control of attention is explored in collaboration with B. Wyble. The eSTST model produces predictions that are tested with a series of 18 experiments aimed at exploring the competitive regulation of attention in several domains, including single streams of characters, spatial arrays, repetitions, words, pictures, and sentence reading. Computational modeling of the sort described here makes potentially informative links with disorders of attention because the model relates behavioral phenomena to neurally plausible structures. As a result, the model's account of attentional control can be related to anatomically specified structures in neuroimaging studies, as well as to temporally specified brainwave signals in EEG studies. The attentional processes and their limits that we propose to study have broad importance for many practical situations in modern life in which the visual environment changes rapidly, including city driving, air traffic control, and response to medical emergencies.

描述（由申请人提供）：该提案的总体目标是理解和模拟那些转瞬即逝的视觉表征形成有意识的、可报告的记忆的过程。当我们寻找某样东西时，当我们的眼睛注视着目标时，我们可以出人意料地快速找到目标。这是通过使用快速串行视觉呈现 (RSVP) 以每秒 10 个甚至 20 个的速率一个接一个地呈现项目来显示的。此类针对字母、单词或图片序列的实验表明，第一个目标很容易被发现，但在接下来的半秒（500 毫秒）内很难发现第二个目标，这种效应称为注意力瞬脱。在这 500 毫秒内，会发生一系列复杂的过程，从检测到第一个目标开始，然后是短暂的增强注意力（称为瞬时注意力），它会迅速将注意力资源吸引到目标的位置。在最初的注意力爆发之后，会发生眨眼，使得此后 200-500 毫秒内很难在视野中的任何地方发现新目标。令人惊讶的是，通过呈现不间断的目标流，这种眨眼现象被完全消除了。因此，一串目标可以保持注意力集中，但如果目标之间存在间隙，注意力就会被关闭。这种效应表明，自上而下和自下而上的注意力控制之间存在着斗争。为了理解这些动态，我们与 B. Wyble 合作探索了注意力时间控制的计算模型。 eSTST 模型产生的预测经过一系列 18 个实验的测试，旨在探索多个领域的注意力竞争性调节，包括单个字符流、空间阵列、重复、单词、图片和句子阅读。这里描述的那种计算模型与注意力障碍建立了潜在的信息联系，因为该模型将行为现象与神经上合理的结构联系起来。因此，该模型对注意力控制的解释可以与神经影像学研究中的解剖学特定结构以及脑电图研究中的时间特定脑电波信号相关。我们建议研究的注意力过程及其局限性对于视觉环境快速变化的现代生活中的许多实际情况具有广泛的重要性，包括城市驾驶、空中交通管制和对医疗紧急情况的响应。