Neural circuits and representational hierarchies supporting human auditory-verbal memory

支持人类听觉语言记忆的神经回路和表征层次结构

• 批准号: RGPIN-2022-04985
• 负责人: Buchsbaum, Bradley
• 金额: $ 2.84万
• 依托单位: University of Toronto
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2022
• 资助国家: 加拿大
• 起止时间: 2022-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=764248
• 关键词: Neural; circuits; representational; hierarchies; supporting

The human brain frequently faces a difficult problem: it receives complex, structured information from the environment and must maintain this information "in mind" for some time before making an appropriate decision or action. The structured information that humans must maintain is often linguistic, arriving from the environment in the form of spoken speech or visual signs. Psychologists and neuroscientists have long been attempting to understand how the brain "breaks down" linguistic information, and then builds it up again into the internal mental structures that can be easily and conveniently manipulated to facilitate thought, action, and other complex mental activities such as verbal working memory. Decades of research on verbal working memory has generally focused on the brain's utilization of the human speech system as a way to temporarily store linguistic information-and the "phonological code" has generally been assumed to be the dominant format for such short-term storage. However, recent research suggests that the working memory system is highly flexible and opportunistic, making use of numerous kinds of "internal codes", distributed throughout the brain, to allow for the manipulation of different types of information in different contexts. By examining distributed patterns of neural activity in different brain regions and relating such patterns to systematic variations in stimulus features, we can use functional magnetic resonance imaging (fMRI) to probe the representations that are active during verbal working memory. In the current program of research, we examine how verbal working memory has the capacity to represent different aspects of human speech, from relatively "low-level" acoustic-phonetic features, to variations in pitch that contribute to prosody, and semantic features that are important to word meaning but are largely unrelated to phonology. We propose three main experiments, each sharing a basic form in which (1) a set of syllables or words are presented to a participant during fMRI scanning, (2) the participant must silently rehearse the words/syllables in mind over a delay, and, finally, (3) recall the items as best they can in serial order. By varying the inter-relationships (phonological, semantic, prosodic) among the word lists across trials, we can use statistical methods to infer which brain regions support the maintenance of different feature representations. We hypothesize that, even though the phonological code is indeed critical for verbal working memory as prior research has shown, that other components of the hierarchical set of features that comprise speech are also important for maintaining information "online". Taken together, this work will greatly expand our understanding on the multidimensional nature of verbal working memory, and the brain systems and networks that support it.

人类大脑经常面临一个难题：它从环境中接收复杂的结构化信息，并且必须在做出适当的决定或行动之前将这些信息“记住”一段时间。人类必须保持的结构化信息通常是语言的，以口头语言或视觉符号的形式来自环境。长期以来，心理学家和神经科学家一直试图了解大脑如何“分解”语言信息，然后将其重新构建到内部心理结构中，这些结构可以轻松方便地进行操作，以促进思维，行动和其他复杂的心理活动，如言语工作记忆。 几十年来对语言工作记忆的研究主要集中在大脑利用人类的语音系统作为暂时存储语言信息的一种方式，而“语音代码”通常被认为是这种短期存储的主要形式。然而，最近的研究表明，工作记忆系统是高度灵活和机会主义的，它利用分布在整个大脑中的多种“内部代码”，允许在不同的环境中处理不同类型的信息。通过检查不同脑区神经活动的分布模式，并将这些模式与刺激特征的系统变化联系起来，我们可以使用功能性磁共振成像（fMRI）来探测言语工作记忆中活跃的表征。在当前的研究计划中，我们研究了言语工作记忆如何能够代表人类语音的不同方面，从相对“低水平”的声学语音特征，到有助于韵律的音高变化，以及对词义很重要但与音韵学无关的语义特征。我们提出了三个主要的实验，每个实验都有一个基本的形式，即（1）在功能磁共振成像扫描过程中向参与者呈现一组音节或单词，（2）参与者必须在一段时间内在脑海中默默地排练单词/音节，最后，（3）以序列顺序尽可能地回忆项目。通过改变试验中单词列表之间的相互关系（语音，语义，韵律），我们可以使用统计方法来推断哪些大脑区域支持不同特征表征的维护。我们假设，即使语音代码确实是至关重要的口头工作记忆先前的研究表明，包括语音的功能的层次设置的其他组件也是重要的，以保持信息“在线”。总的来说，这项工作将大大扩展我们对语言工作记忆的多维性质以及支持它的大脑系统和网络的理解。