Functional organization of the superior temporal gyrus for speech perception

描述（由申请人提供）：言语感知是听者听到和解释语言声音的过程。它是由高度可变和复杂的声学信号到语音表示的独特映射定义的。一个趋同的研究表明，人类的后上级颞叶皮层专门处理语音。虽然语言信息是局部的，但提取语言信息的基本神经机制完全不清楚。我们提出了一种创新的方法学方法，使用定制的颅内高密度电极阵列直接记录清醒的、行为正常的接受癫痫手术临床评价的受试者的后颞叶皮层的详细神经活动。这种方法提供了最高的空间和时间分辨率，从而克服了非侵入性成像方法的许多限制。本研究定位于听觉神经科学和语言学领域之间的关键接口。我们 先前的结果表明，语音的皮层表示表现出重要的非线性，其对应于声学参数上的感知边界（Chang等人，NatureNeuroscience 2010）。基于这些发现，我们提出了实验来确定：1）后颞回的声学和语音特征的功能组织，2）分布式，人口水平的编码的紧急语音表示，和3）相关的“鲁棒”的语音表示在一个多说话人的听力环境的背景下。这些实验的结果将对神经语言学领域和更广泛的感官知觉和认知研究产生重大影响。

DESCRIPTION (provided by applicant): Speech perception is the process by which listeners hear and interpret the sounds of language. It is defined by the unique mapping of a highly variable and complex acoustic signal to a phonetic representation. A convergence of studies has implicated the human posterior superior temporal cortex for the specialized processing of speech sounds. Although localized, the basic neural mechanisms by which linguistic information is extracted are entirely unclear. We propose an innovative methodological approach using customized intracranial high-density electrode arrays to record detailed neural activity directly from the posterior temporal cortex in awake, behaving subjects undergoing clinical evaluation for epilepsy surgery. This method offers the highest possible spatial and temporal resolution, thereby overcoming many limitations of non-invasive imaging approaches. This research is positioned at a critical interface between the fields of auditory neuroscience and linguistics. Our previous results demonstrate that cortical representation of speech sounds manifest important non-linearities that correspond to perceptual boundaries over acoustic parameters (Chang et al, Nature Neuroscience 2010). Building on these findings, we propose experiments to determine: 1) the functional organization of the posterior temporal gyrus for acoustic and phonetic features, 2) the distributed, population-level encoding of emergent phonetic representation, and 3) the correlates of "robust" speech representation in the context of a multi-speaker listening environment. The results of these proposed experiments will have significant impact on the field of neurolinguistics and broader research on sensory perception and cognition.