The role of amplitude modulation in perceiving speech and music

幅度调制在感知语音和音乐中的作用

• 批准号: 10728708
• 负责人: Andrew Chang
• 金额: $ 0.25万
• 依托单位: NEW YORK UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-03-01 至 2023-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10728708
• 关键词: Acoustics; Address; Affect; Algorithms; Alzheimer' s Disease; Aphasia; Auditory; Auditory Perception; Auditory area; Behavioral; Brain; Brain region; Cognition; Communication; Data; Frequencies; Functional Magnetic Resonance Imaging; Goals; Human; Judgment; Linguistics; Machine Learning; Magnetoencephalography; Measures; Mind; Music; Noise; Participant; Pathway interactions; Perception; Performance; Periodicity; Persons; Process; Psychophysics; Records; Research; Research Personnel; Role; Series; Signal Transduction; Solid; Speech; Speech Perception; Stimulus; Structure; System; Testing; Work; auditory processing; career; cognitive neuroscience; experimental study; individuals with autism spectrum disorder; insight; neural; neuroimaging; neuromechanism; operation; programs; sound; spectral energy; speech processing

Project Summary/Abstract My career goal is to become a leading researcher on cognitive neuroscience, with a special focus on the neural mechanisms underlying auditory perception, including how humans track and perceive the fleeting audi- tory information in speech and music. In this proposal, I outline a research program to investigate the acoustic and neural distinctions between speech and music, two specialized forms of auditory signals that are closely tied to the human mind. Despite our increasingly rich understanding of the perceptual and neural mechanisms for processing speech or music, surprisingly little is known about why and how they are treated as different au- ditory signals by the human mind and brain in the first place. Investigating these distinctions is foundational for a thorough understanding of how acoustic waveforms are transformed into meaningful information. The work will provide a more solid basis for understanding cognition and communication as well as treating people with communicative deficits, such as people with autism, Alzheimer's disease, and aphasia. I hypothesize that the temporal structure reflected in the amplitude modulation (AM) of speech and music signals is a critical distinctive feature for the brain and engages to different processing pathways, as speech and music are known to have distinct AM rates. A series of studies, combining psychophysics, MEG (magne- toencephalography), fMRI (functional magnetic resonance imaging), and machine learning approaches, will use stimuli with AM rates across the modulation frequency ranges of speech and music to address this topic at the computational (the goals), algorithmic (the representations and operations), and implementational (neural mechanism) levels. (1) Does the AM rate of a sound affect whether it will be perceived as speech or music? (2) Does the AM rate of a stimulus optimize speech and music perceptual performance at different frequencies? (3) What are the underlying neural mechanisms and the associated brain regions implementing the differentiation of speech and music? Aim 1 investigates whether the AM rate of a sound conditions it to be processed as speech or music. By manipulating the AM rate of noise-vocoded speech and music recordings, I hypothesize that the sounds with slower or faster AM rates will likely to be perceived as music or speech, respectively, the perceptual judgment will be biased by the higher or lower spectral energy of neural oscillatory activity (meas- ured by MEG) while listening to the sounds, respectively, and the associated brain regions will be revealed by fMRI with machine learning decoding approaches. Aim 2 investigates whether the AM rate of stimuli optimizes speech and music perceptual performances at different rates. I hypothesize that the music perceptual perfor- mance is optimal at slower AM rates while the speech perceptual performance is optimal at faster AM rates, and the neural oscillatory entrainment at lower or higher frequency band has domain-specific function facilitat- ing speech or music perceptual performance.

项目概要/摘要 我的职业目标是成为认知神经科学领域的领先研究员，特别关注 听觉感知背后的神经机制，包括人类如何跟踪和感知转瞬即逝的听觉 言语和音乐中的理论信息。在这个提案中，我概述了一个研究计划来调查声学 以及语音和音乐之间的神经区别，这两种特殊形式的听觉信号密切相关 与人类的思想联系在一起。尽管我们对感知和神经机制的了解越来越丰富 对于处理语音或音乐，令人惊讶的是，人们对为什么以及如何将它们视为不同的au-知之甚少。 首先，人类思想和大脑发出的历史信号。研究这些区别是以下研究的基础 全面了解声波波形如何转化为有意义的信息。工作 将为理解认知和沟通以及对待人们提供更坚实的基础 沟通障碍，例如患有自闭症、阿尔茨海默病和失语症的人。 我假设语音和音乐的幅度调制（AM）反映了时间结构 信号是大脑的一个重要的独特特征，参与不同的处理途径，例如语音 众所周知，音乐具有不同的 AM 速率。一系列研究结合了心理物理学、MEG（磁力 脑成像）、fMRI（功能磁共振成像）和机器学习方法，将 在语音和音乐的调制频率范围内使用 AM 速率的刺激来解决这个主题 计算（目标）、算法（表示和操作）和实现（神经 机制）水平。 (1) 声音的 AM 速率是否会影响它被感知为语音还是音乐？ (2) 刺激的 AM 速率是否可以优化不同频率下的语音和音乐感知性能？ (3) 实现分化的潜在神经机制和相关大脑区域是什么 言语和音乐？目标 1 研究声音的 AM 速率是否将其处理为 演讲或音乐。通过操纵噪声声码语音和音乐录音的 AM 速率，我假设 调幅速率较慢或较快的声音可能分别被视为音乐或语音， 感知判断会因神经振荡活动的较高或较低光谱能量而产生偏差（测量 分别通过 MEG 驱动），同时聆听声音，相关的大脑区域将通过 功能磁共振成像与机器学习解码方法。目标 2 研究刺激的 AM 速率是否优化 言语和音乐感知表现以不同的速率。我假设音乐感知表现 曼斯在较慢的 AM 速率下最佳，而语音感知性能在较快的 AM 速率下最佳， 并且较低或较高频带的神经振荡夹带具有特定领域的功能促进 言语或音乐的感知表现。