Dual Speech Coordination: fMRI Investigations of a New Neuroarchitectural Model of Speech Production

语言是人类交流的一种普遍形式，然而数以百万计的美国人，无论老少，在表达语言方面都有困难，这影响了他们的福祉和生活。语言障碍是一种常见的与大脑有关的问题，它可以由广泛的遗传、损伤或退行性大脑问题引起。帮助有这种障碍的人可能很困难，因为语言是一种极其复杂的大脑能力，往往很难找到问题的根源，从而进行治疗。本研究的目标是了解语音产生系统的几个关键部分。这项研究使用先进的脑成像技术来测量健康人在执行几种不同的语音生成任务时的大脑活动模式。这些研究使我们能够生成健康大脑系统的功能“地图”，使我们能够流利地说话。这些地图可以与大脑疾病导致的语言产生问题的人的大脑变化模式相关联，使我们能够确定问题的根源，并使开发恢复语言流畅性的疗法成为可能。了解大脑如何将思想转化为语言的详细机制——这个项目的总体目标——不仅使我们能够改进对语言产生缺陷的诊断和治疗，而且还为合成语音和语言学习教育技术的发展带来了新的见解。这个项目的具体目标是评估一种假设，即语言产生计划被分为两个可分离的过程，由两个不同的大脑系统控制，一个涉及与音高相关的发声协调（旋律或歌唱般的语音质量，称为语调或韵律，在对婴儿说话时经常被强调），另一个涉及清晰的语言协调（形成和排序音素，音节，单词和短语）。这一观点的实际重要性在于，协调清晰的言语的能力可能取决于产生与音调相关的话语韵律的能力，而韵律是言语计划的支撑。这与将韵律视为附加能力而非基础的典型思维不同。这种新的观点，如果得到研究证实，可能会从根本上改变我们对语言产生方式的理解。为了研究这个问题，我们使用脑成像技术（功能性磁共振成像或fMRI）来绘制与语音产生有关的神经回路，使用任务，使我们能够分别测量从语音和音节语音内容的发音中产生与音高相关的韵律轮廓的能力。我们期望这些实验能够揭示支持这两种语言产生能力的可区分的神经回路，从而为临床评估语言障碍的诊断工具和更好的语言产生的计算/合成模型的开发打开大门。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Speech is a universal form of human communication, and yet millions of Americans, young and old, have difficulties in producing speech, affecting their well-being and livelihood. Speech impairment is a common brain-related problem that can be caused by a wide range of genetic, injury-related, or degenerative brain problems. Helping people with such impairments can be difficult because speech is an incredibly complicated brain ability, often making it hard to pinpoint the source of the problem so that it can be treated. The goal of the proposed studies is to understand several key parts of the speech production system. The research uses advanced brain imaging techniques to measure the brain’s activity pattern while healthy people are performing several different speech production tasks. These investigations allow us to generate a functional “map” of the systems of the healthy brain that enable us to speak fluently. These maps can then be related to patterns of brain changes in people with speech production problems due to brain disease, allowing us to identify the source or sources of the problem, and making it possible to develop therapies to recover speech fluency. Understanding the detailed mechanisms of how the brain transforms thoughts into speech—the overarching goal of this project—not only allows us to improve diagnosis and treatment of speech production deficits but also leads to new insights in the development of technologies for synthetic speech and language learning education. The specific goal of this project to evaluate the hypothesis that speech production planning is divided into two separable processes controlled by two distinct brain systems, one involving the coordination of pitch-related vocalization (the melodic or sing-song like quality of speech known as intonation or prosody that is often emphasized in speaking to babies) and the other involving the coordination of articulate speech (forming and sequencing phonemes, syllables, words and phrases). The practical importance of this idea is that the ability to coordinate articulate speech may be dependent on the ability to generate the pitch-related prosody of an utterance, which acts as a scaffold for speech planning. This differs from typical thinking in which prosody is an add-on ability rather than a foundation. This new perspective, if borne out by research, may fundamentally change our understanding of how speech is produced. To study this question, we use a brain imaging technique (functional magnetic resonance imaging or fMRI) to map neural circuits involved in speech production using tasks that allow us to separately measure the ability to generate pitch-related prosodic contours from the articulation of phonetic and syllabic speech content. We expect these experiments to reveal distinguishable neural circuits supporting these two speech production abilities, which opens the door to the development of diagnostic tools for evaluating speech disabilities in the clinic and better computational/synthetic models of speech production.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.