Studying semantic processing during language comprehension in humans at the single-cellular level

在单细胞水平上研究人类语言理解过程中的语义处理

• 批准号: 10591471
• 负责人: Ziv Williams
• 金额: $ 54.89万
• 依托单位: MASSACHUSETTS GENERAL HOSPITAL
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-03-15 至 2027-02-28
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10591471
• 关键词: Action Potentials; Algorithms; Animal Model; Animals; Aphasia; Area; Brain; Cells; Chronic; Cognition; Communication; Communication impairment; Communities; Complex; Computational Technique; Computer Models; Data; Dependence; Derivation procedure; Disease; Dyslexia; Electrodes; FDA approved; Goals; Grain; Higher Order Chromatin Structure; Human; Impairment; Implant; Individual; Institution; Institutional Review Boards; Language; Language Disorders; Left; Letters; Linguistics; Maps; Microelectrodes; Modeling; Neurons; Participant; Population; Prefrontal Cortex; Process; Psychological Techniques; Research; Resolution; Semantics; Speech; Speech Sound; Structure; Techniques; Temporal Lobe; Testing; Time; Tissues; Validation; autism spectrum disorder; cognitive process; combinatorial; frontal lobe; insight; language comprehension; neural; neurophysiology; novel; prospective; response; semantic processing; syntax; temporal measurement; theories

Studying semantic processing during language comprehension in humans at the single-cellular level Humans are capable of communicating extraordinarily rich and nuanced meanings through language. This capacity surpasses that of all other animal species. Yet, despite a growing understanding of the network of brain areas that supports semantic processing, the precise derivation of linguistic meaning in neural tissue at the cellular-level and over the temporal scales of action potentials remains largely unknown. Over the past several years, our collaborative group has developed unique approaches that have allowed us to record from individual neurons in the language-dominant prefrontal and temporal cortex in participants performing structured linguistic tasks. Here, we will use this extraordinarily rare opportunity together with our group’s unique combined expertise in linguistic theory, human intraoperative neurophysiology, single-neuronal recordings and computational modeling in order to study in detail, for the first time, how semantic information is encoded at a cellular-level during speech and the degree to which single neurons respond selectivity to specific semantic domains, whether semantic information can be robustly decoded from neural activity, the extent to which it can be generalized across linguistic materials and the process by which these neuronal activities maps onto the fine-grained semantic relationships between individual words. Using structured linguistic manipulations and violations and by employing information-theoretic techniques, we will also crucially examine whether certain neurons in humans respond selectively to linguistic compared to non-linguistic information, how they engaged in lexico-semantic or syntactic processes, how they track the real-time context-dependent inferred meanings of individual words during speech and how these mixed representations and computations are distributed within frontal and temporal cortical populations in the language-dominant hemisphere. Taken together, this novel cross-institutional, inter-disciplinary collaborative effort promises to provide a fundamental new platform by which to begin studying the basic cellular mechanisms that underlie human language and unprecedented new insight into the cellular-level processing and representation of linguistic meaning during language comprehension in humans.

研究人类语言理解过程中的语义处理 单细胞水平 人类能够通过语言传达非常丰富和微妙的含义。这 能力超过了所有其他动物物种。然而，尽管人们对网络的了解越来越多， 支持语义处理的大脑区域，即神经组织中语言意义的精确推导， 细胞水平和动作电位的时间尺度在很大程度上仍然是未知的。过去 几年来，我们的合作小组已经开发出独特的方法，使我们能够记录从 参与者语言主导的前额叶和颞叶皮层中的单个神经元 结构化的语言任务。在这里，我们将利用这个非常难得的机会，与我们集团的 在语言理论、人类术中神经生理学、单神经元 录音和计算建模，以详细研究，第一次，如何语义 信息在语音过程中是在细胞水平上编码的，单个神经元的反应程度 对特定语义域的选择性，语义信息是否可以从神经网络中鲁棒地解码， 活动，它可以在语言材料中推广的程度以及这些活动的过程 神经元活动映射到各个单词之间的细粒度语义关系上。使用 结构化的语言操纵和侵犯，并采用信息理论的技术，我们 还将研究人类的某些神经元是否对语言有选择性地做出反应， 非语言信息，他们如何参与词汇语义或句法过程，他们如何跟踪 实时上下文相关的推断意义的个别单词在讲话期间，以及这些如何混合 表示和计算分布在额叶和颞叶皮层人群中， 语言优势半球总之，这种新颖的跨机构，跨学科 合作的努力有望提供一个新的基础平台，通过它开始研究基本的 人类语言的细胞机制和对细胞水平前所未有的新见解 人类在语言理解过程中对语言意义的加工和表征。