Auditory nerve encoding of speech with varied acoustics

听觉神经编码不同声学的语音

• 批准号: 6406249
• 负责人: HANNA E STEVENS
• 金额: $ 2.48万
• 依托单位: UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN
• 依托单位国家: 美国
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-06-11 至
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/6406249
• 关键词: acoustic nerve; auditory stimulus; chinchilla; electrophysiology; neural information processing; phonology; predoctoral investigator; speech

DESCRIPTION (provided by applicant): The broad aims of this project are to understand normal auditory processing of speech sounds, information that can be used to understand and provide rehabilitation for deficits in hearing and speech comprehension. Responses to speech sounds by the auditory nerve (AN) of an animal model, the chinchilla, demonstrate how information about speech sounds is encoded by the mammalian auditory system. A peripheral auditory system like most mammals and the ability of the chinchilla to demonstrate phoneme perception and categorization make it an excellent animal model. Unique temporal patterns of encoding in the AN will be identified for each phoneme across variable acoustics. Co-articulation with a vowel and the absence of normal voicing in whispered speech both alter the acoustics of a consonant. These temporal patterns will demonstrate the role that the auditory system plays in accurate human perception of one phoneme across acoustic profiles and what information is necessary for hearing-impaired people to understand speech. Responses of single AN fibers to normally-voiced and whispered speech will be recorded and analyzed to demonstrate phoneme-specific patterns of response. The specific questions to be asked for each aim: 1) What temporal patterns are characteristic of the AN response to each phoneme /d/ and /t/ across different acoustic profiles? This will be addressed by using as the stimuli four syllables /da/, /ta/, /dae/, and /tae/, from two speakers produced in both whispered and normal voicing. The AN responses to the eight tokens of each consonant will be compared using a global average peri-stimulus time (PST) histograms and local average PST histograms. 2) Are these temporal patterns unique to /d/ and /t/ or unique to the categories of voiced and unvoiced consonants? Four tokens of /b/ and /p/ will be used as stimuli to compare the AN response to consonants within and between voicing categories.

描述（由申请人提供）：本项目的主要目标是 理解语音的正常听觉处理，可以被 用于理解和提供听力和语言缺陷的康复 理解力听觉神经（AN）对语音的反应 动物模型，栗鼠，演示了语音信息是如何 由哺乳动物的听觉系统编码一个外围听觉系统， 大多数哺乳动物和灰鼠的能力，以证明音素的知觉 和分类使其成为极好的动物模型。独特的时间模式 AN中的编码将被识别为跨变量的每个音素 声学.与元音的共发音和正常发音的缺失， 耳语都改变了辅音的声学效果。这些时间 模式将证明听觉系统在准确的 人类对一个音素在声学轮廓上的感知以及什么信息 这是听力障碍者理解语言所必需的。响应单 将记录和分析正常发声和低声说话的AN纤维 来展示音素特定的反应模式。具体问题， 对于每个目标都要问：1）AN的特征是什么时间模式 在不同的声学轮廓上对每个音素/d/和/t/的响应？这 将通过使用四个音节/da/、/ta/、/dae/和 /tae/，从两个扬声器产生的耳语和正常的发声。的AN 对每个辅音的八个标记的响应将使用全局 平均刺激周围时间（PST）直方图和局部平均PST直方图。 2)这些时间模式是/d/和/t/独有的，还是类别独有的 清辅音和浊辅音的区别吗四个标记/B/和/p/将用作 刺激，以比较AN反应的辅音内部和之间的发声 类别