Neural dynamics of speech encoding and maintenance during phonological working memory

项目总结 语音工作记忆是维持和操纵语音表征的能力 声音暂时存在于工作记忆中。人们认为它在语言的许多方面都起着至关重要的作用 发展，如词汇学习、句子处理和识字技能的获得，以及缺陷 语音工作记忆是许多发育性语言障碍的诊断。经典 语音工作记忆的理论框架假设信息被保存在专门的工作中 记忆存储缓冲，独立于语言感知皮质；尽管较新的认知模型支持 一种观点，在这种观点中，同样的皮质底物也负责感知语音信息 支持在工作记忆延迟期间维护这些神经表示。事实上，新近的人类 神经成像研究表明语音工作记忆负荷依赖于言语知觉的招募 大脑皮质在各种任务中的激活，特别是左额叶上回的激活已被证明 在语音工作记忆的临床评估中，与扫描仪外的表现相关。然而， 因为这些研究不能在时间上区分语音编码过程中的激活和 编码后语音表征的持续保持，我们不知道这些中的哪一个 将过程归因于观察到的与行为的相关性。因此，从理论上讲， 这些发现，以及它们的临床解释，在很大程度上取决于解决语音编码和 工作记忆在时间和空间上的维持过程。我们建议研究神经动力学。 使用脑电的语音工作记忆，它允许同时 通过放置电极以毫米和毫秒分辨率记录多个离散的神经群 直接在清醒的人的大脑皮质表面上，表现为接受手术的人类患者 医学上难治性癫痫。这项技术将使我们能够直接观察言语之间的关系 语音工作记忆过程中的编码与维持。通过三个具体目标，我们将：(1) 确定大脑皮层语音工作记忆中神经激活的时间进程，(2) 确定在编码过程中表示语音信息的相同感知系统是否也 在工作记忆过程中主动保持这些表征，以及(3)确定知觉言语是否 在工作内存维护期间，系统参与更广泛的网络级动态。通过实现 这些目标，这个项目将促进我们对感知语音系统在 语音工作记忆。最终，这些发现可以被用来为角色的理论基础提供信息 语音工作记忆在语言发展中的作用及其对临床特征的影响 以及修复语音工作记忆缺陷。

PROJECT SUMMARY Phonological working memory is the capacity to maintain and manipulate representations of speech sounds temporarily in working memory. It is believed to play a critical role in many aspects of language development, such as vocabulary learning, sentence processing, and acquisition of literacy skills, and deficits of phonological working memory are diagnostic of many developmental language disorders. The classic theoretical framework for phonological working memory posits that information is held in a dedicated working memory storage buffer, separate from speech perception cortex; although more recent cognitive models favor a view in which the same cortical substrate responsible for perception of phonological information also supports maintenance of those neural representations during working memory delay. Indeed, recent human neuroimaging studies show phonological working memory load-dependent recruitment of speech-perception cortex during a variety of tasks and in particular, activation in left superior temporal gyrus has been shown to correlate with out-of-scanner performance on clinical assessments of phonological working memory. However, because these studies are not able to temporally distinguish activation during phonological encoding from persistent maintenance of phonological representations after encoding, we do not know which of these processes to attribute to the observed correlations with behavior. Therefore, the theoretical implications of these findings, as well as their clinical interpretations, critically depend on resolving phonological encoding and working memory maintenance processes in time and space. We propose to investigate the neural dynamics of phonological working memory using intracranial electrocorticography, which allows for the simultaneous recording of many discrete neural populations with millimeter- and millisecond-resolution by placing electrodes directly on the cortical surface of the brains of awake, behaving human patients undergoing surgery for medically intractable epilepsy. This technique will allow us to observe directly the relationship between speech encoding and maintenance during phonological working memory. Through three specific aims we will (1) determine the time course of neural activation during phonological working memory across the cortex, (2) determine whether the same perceptual systems that represent phonological information during encoding also actively maintain those representations during working memory, and (3) determine whether perceptual speech systems participate in broader network-level dynamics during working memory maintenance. By achieving these aims, this project will advance our understanding of the role perceptual speech systems play during phonological working memory. Ultimately, such findings can be used to inform the theoretical basis for the role of phonological working memory in language development and has implications for the clinical characterization and remediation of phonological working memory deficits.