Use of fMRI to inform the components of a cognitive architecture

描述(申请人提供)：本项目的目标是使用功能磁共振脑成像来了解复杂认知能力的组织和获得。能够模拟复杂认知能力的ACT-R认知架构的组成部分已经被映射到不同的大脑区域。已经开发出一种方法来获取这些成分的活动，并预测从功能磁共振成像中在这些区域获得的大胆反应。最近的发展也使这种测试能够在单个部件的时间高度变化性的情况下应用。这项研究旨在进一步发展ACT-R理论各组成部分之间的映射，并扩展这项工作以理解复杂任务的结构。提出了三类实验。首先，实验将测试与陈述性提取相关的前额叶区域、与心理表征相关的顶叶区域以及与控制相关的前扣带回区域的预测。第二组实验旨在扩展和修正ACT-R理论。他们将研究前额叶区域的反应是否可以指导ACT-R中元认知成分的发展，以及是否应该修改ACT-T的背尾状核理论，以包括强化学习的影响。第三类实验将测试分解假设--大任务中的过程由与小任务中相同的组件组成。将对一个复杂的雷达屏幕任务进行调查，看看该任务中的激活模式是否可以从ACT-R理论的单个组件的行为中预测出来。还将对其进行研究，以确定认知架构进一步发展的方向。除了促进对复杂认知的理解外，这项研究还将有助于开发利用功能磁共振研究来研究更高水平认知功能的方法。将fMRI脑成像映射到更高水平的认知功能对于许多应用非常重要，包括与健康相关的工作，如了解认知功能障碍的基础。例如，许多自闭症理论将其与不同大脑区域为智力目标服务的协调有关。公共卫生相关性将fMRI脑成像映射到更高级别的认知功能对于许多应用具有重要意义，包括与健康相关的工作，如了解认知功能障碍的基础。例如，许多自闭症理论将其与不同大脑区域为智力目标服务的协调有关。

DESCRIPTION (provided by applicant): The goal of this project is to use fMRI brain imaging to understand the organization and acquisition of complex cognitive abilities. Components of the ACT-R cognitive architecture, which is capable of modeling complex cognitive abilities, have been mapped on different brain regions. A methodology has been developed for taking the activities of these components and making predictions for the BOLD response obtained in these areas from fMRI imaging. Recent developments have also enabled such tests to be applied in the face of high variability in the timing of individual components. The proposed research is intended to further develop the mapping between components of the ACT-R theory and to extend the work to understand the structure of complex tasks. Three classes of experiments are proposed. First, experiments will be performed to test predictions about a prefrontal region associated with declarative retrieval, a parietal region associated with mental representation, and an anterior cingulate region associated with control. The second group of experiments are designed to extend and modify the ACT-R theory. They will examine whether the response of an anterior prefrontal region can guide the development of a metacognitive component in ACT-R and whether the ACT-T theory of the dorsal caudate should be amended to include effects of reinforcement learning. The third class of experiments will test the decomposition hypothesis -- that the processes in large task consist of the same components as are revealed in smaller tasks. A complex radar-screen task will be investigated to see if the activation patterns in this task can be predicted from the behavior of individual components of the ACT-R theory. It will also be studied to identify further directions for the development of the cognitive architecture. In addition to advancing understanding of complex cognition this research will serve to develop the methodology for using fMRI research to study higher-level cognitive functioning. Advancing the mapping of fMRI brain imaging to higher-level cognitive function is important for many applications including health-related efforts such as understanding the basis of cognitive dysfunctions. For instance, many theories of autism relate it to the coordination of different brain regions in service of intellectual goals. PUBLIC HEALTH RELEVANCE Advancing the mapping of fMRI brain imaging to higher-level cognitive function is important for many applications including health-related efforts such as understanding the basis of cognitive dysfunctions. For instance, many theories of autism relate it to the coordination of different brain regions in service of intellectual goals.