言语识别中感觉运动整合的脑机制以及听觉老年化和音乐训练的影响作用

Investigation of brain mechanisms underlying speech recognition and comprehension in complex and noisy environments is an important approach to enrich the understanding of human perception and cognition. One important cross-modal mechanism to facilitate speech recognition under adverse listening conditions is sensorimotor integration. It is hypothesized that articulatory predictions from speech motoric areas can top-down modulate speech representations in auditory cortices, particularly at uncertain and noisy conditions, but the neural mechanism underlying speech sensorimotor integration is yet largely unclear. In addition, auditory aging and long-term musical training are associated with impaired and facilitated speech recognition in noisy environments, respectively. We propose that older listeners may increase the dependence on sensorimotor integration to compensate aging-related declines in auditory processing; while musicians may show enhanced sensorimotor integration that contributes to their improvements in speech recognition at noisy conditions. This project will combine psychophysical measurements, magnetoencephalography (MEG) and functional magnetic resonance imaging (fMRI) to systematically investigate the neural mechanisms underlying sensorimotor integration of speech perception, as well as the separate effects of and interactions between auditory aging and long-term musical training on sensorimotor integration. This project will focus on the following aspects of sensorimotor integration process: the anatomical and functional connectivities as well as the oscillatory synchronization mechanisms in sensorimotor integration network; the recruitment and categorical phoneme-specific representations in speech motor and auditory areas; and the responding pattern of sensorimotor integration to the level of masking. Findings from this project will not only advance the understanding of the "cocktail-party problem" and the multi-modal integration mechanisms in complex cognitive processes, but also be valuable in applications in speech rehabilitation and training programs for senior populations and more.

认识复杂场景下言语识别和理解的脑机制对理解大脑感知觉和认知功能具有重要意义。一种能在嘈杂环境中促进言语识别但对其认识尚浅的机制是语音感觉运动整合：听者语音运动区的发音预期对听皮层语音编码的下行调节。同时，听觉老年化和长期音乐训练分别降低和增强了嘈杂环境中的言语识别能力。本项目首次提出，老年人可能更依赖语音感觉运动整合以代偿受损听觉加工，而音乐训练可增强感觉运动整合以促进言语分辨。本项目将结合心理物理学、脑磁和磁共振成像系统性研究语音感觉运动整合过程的神经基础和作用机理，同时考察老年化和音乐训练在以下方面的分别影响和协同作用：语音感觉运动整合环路的结构、功能连接和同步振荡机制，语音运动区和听皮层的激活和音素范畴特异性表征模式，以及感觉运动整合对掩蔽程度的动态响应模式。本项目研究成果将加深对“鸡尾酒会问题”和复杂认知过程中多模态信息整合机制的理解，并在老年人语音康复训练等方面具有巨大应用价值。

认识复杂场景下言语识别和理解的脑机制对理解大脑感知觉和认知功能具有重要意义。一种能在嘈杂环境中促进言语识别但对其认识尚浅的机制是语音感觉运动整合：听者发音运动区的发音预测对听皮层语音编码的下行调节。听觉老化和音乐训练经验被发现可分别降低和增强嘈杂环境中的言语识别能力。本项目结合心理物理学、功能磁共振成像fMRI与扩散张量成像DTI和经颅磁刺激TMS技术，共开展9项研究，发表包括PNAS和Cerebral Cortex在内的SCI Q1论文3篇，Q2论文1篇（项目负责人均为通讯作者）。在学术价值方面，本项目首先在年轻被试中深入探索了在嘈杂环境中促进言语识别的跨通道感觉运动整合机制的神经基础和作用机理，包括利用脑成像元分析技术揭示了语音感觉运动整合的神经环路（Frontiers in Neuroscience, 2018），利用TMS技术揭示了左侧喉部运动区因果性参与声调感知的神经机制以及利用fMRI技术揭示了发音运动信息促进语音感觉运动整合环路的局部神经表征和动态功能连接（两项成果待发表）。更重要的是，本项目加深了对老化和音乐训练经验如何分别和交互影响嘈杂环境中言语加工能力的神经机制的理解。包括利用fMRI技术在神经激活模式、局部神经表征及功能和结构连接层面均发现，老年人更依赖语音感觉运动整合以代偿受损的听觉言语感知，而音乐训练经验可增强感觉运动整合功能以促进噪音下言语感知（PNAS, 2017; Cerebral Cortex已接收）。同时，通过对比年轻和老年的音乐家和非音乐家群体，本项目在行为和脑成像层面均证实，音乐训练经验可增强语音感觉运动整合及高级认知功能以对抗和延缓老化引发的噪音下言语感知能力的下降（Ear & Hearing, in press; 两项成果数据整理中）。本项目研究成果不仅加深了对“鸡尾酒会问题”和复杂认知过程中多模态信息整合机制的理解，在应用价值上，本研究提示综合性的音乐训练及感觉运动整合的靶向训练或刺激干预对改善老年群体以及言语障碍人群的言语识别和加工能力具有重要的临床价值。

内容获取失败，请点击重试