Neurocognitive mechanisms of task representation reorganization during task learning

任务学习过程中任务表征重组的神经认知机制

• 批准号: 10561485
• 负责人: Jiefeng Jiang
• 金额: $ 54.31万
• 依托单位: UNIVERSITY OF IOWA
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-16 至 2027-05-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10561485
• 关键词: Adaptive Behaviors; Affect; Animals; Artificial Intelligence; Attention deficit hyperactivity disorder; Behavior; Behavioral; Brain; Brain imaging; Clinical; Cognitive; Collection; Complex; Data; Disease; Functional Magnetic Resonance Imaging; Goals; Hippocampus (Brain); Human; Human Characteristics; Intelligence; Knowledge; Learning; Link; Maps; Measures; Memory; Mental disorders; Methods; Modeling; Neurocognitive; Participant; Patients; Pattern; Performance; Process; Psyche structure; Quality of life; Research; Response to stimulus physiology; Retrieval; Role; Schizophrenia; Shapes; Specific qualifier value; Statistical Data Interpretation; Structure; Support System; System; Task Performances; Testing; Thinking; Work; base; cognitive control; executive function; experience; experimental study; flexibility; improved; information processing; insight; intelligent agent; learning ability; long term memory; neuroimaging; procedural memory; programs; relating to nervous system; relational memory; response; young adult

Project Summary/Abstract Humans possess extraordinary flexibility in our behavior: Given the same environmental input, we can act differently depending on our goals and the context. For example, when facing the same data, we can process them differently depending on our goals (e.g., visualize the data to obtain a figure, perform statistical analysis to test a prediction, or even delete the data if the goal is to free storage space). To date, research on this topic has focused on how such flexible behavior is implemented via cognitive control, which is a set of cognitive mechanisms supporting goal-directed and top-down modulation on information processing in the brain. In other words, much research has been conducted to study how a task is executed. However, less is known about where such task knowledge is from, that is, the mnemonic mechanisms that encode, reinforce, and generalize the neural representations of task knowledge. Understanding these mechanisms is crucial to fully understand human intelligence, as the remarkable abilities of learning and retaining task knowledge promptly and efficiently distinguish humans from other animals and artificial intelligent agents and make us adaptive to this ever-changing world. Furthermore, filling the knowledge gap of how we learn and remember task knowledge is also key to understand, detect and treat task learning deficits that are common in mental disorders such as schizophrenia and attention-deficit / hyperactivity disorder (ADHD). In this project, we will focus on the hippocampus, a central brain structure for learning and memory. To achieve the objective of uncovering the hippocampal contributions to task learning, six experiments are proposed using a combination of behavioral methods and human functional magnetic resonance imaging. Specifically, Aim 1 will identify hippocampal contribution to constructing a task representation by assembling task information and experiences to build a task model. Aim 2 will identify how the hippocampus encodes a new task representation into a memory network of existing representations. Aim 3 will identify how the hippocampus reshapes existing task representations when they become associated with other tasks via compositional relations. This work is expected to identify how the human hippocampus constructs key content of task representations (Aim 1) and organizes multiple task representations in relation to each other (Aim 2 and 3). These findings will further our understanding of how the hippocampus contributes to task learning, cognitive control and adaptive behavior. This work will also have clinical impact in bridging the gap between hippocampal abnormality and task learning deficits, which were separately observed in mental disorders such as schizophrenia and ADHD. Ultimately, this work will have a broad impact in helping detect and treat task learning deficits.

项目总结/摘要 人类在行为上具有非凡的灵活性：在相同的环境输入下，我们可以采取行动， 根据我们的目标和背景而有所不同。例如，当面对相同的数据时， 它们根据我们的目标而不同（例如，将数据可视化以获得图形，执行统计分析 以测试预测，或者如果目标是释放存储空间，则甚至删除数据）。迄今为止，关于这一主题的研究 他专注于如何通过认知控制来实现这种灵活的行为，这是一套认知控制。 支持大脑中信息处理的目标导向和自上而下调制的机制。换句 换句话说，已经进行了许多研究来研究如何执行任务。然而，人们对 这些任务知识来自何处，即编码、强化和概括的记忆机制 任务知识的神经表征。理解这些机制对于充分理解 人类智力是人类迅速学习和保持任务知识的显著能力， 有效地区分人类与其他动物和人工智能代理，使我们适应这一点， 不断变化的世界此外，填补我们如何学习和记忆任务知识的知识缺口是 也是理解、检测和治疗任务学习缺陷的关键，这些缺陷在精神障碍中很常见， 精神分裂症和注意力缺陷多动障碍（ADHD）。在本项目中，我们将重点关注 海马体，一种负责学习和记忆的大脑中枢结构。为了达到揭露 海马的贡献，任务学习，六个实验提出了使用行为的组合 方法和人类功能性磁共振成像。具体来说，目标1将识别海马 通过组装任务信息和经验来构建任务表示， 任务模型目标2将确定海马体如何将新的任务表征编码到记忆中 现有的网络代表。目标3将确定海马体如何重塑现有任务 当它们通过组合关系与其他任务相关联时，它们会被表示。这项工作是 期望确定人类海马体如何构建任务表征的关键内容（目标1）， 将多个任务表示相互联系起来（目标2和3）。这些发现将进一步推动我们 了解海马体如何有助于任务学习、认知控制和适应行为。 这项工作也将在弥合海马异常和任务学习之间的差距方面产生临床影响 这些缺陷在精神障碍如精神分裂症和ADHD中分别观察到。最终这 工作将在帮助发现和治疗任务学习缺陷方面产生广泛的影响。