Single-neuron population dynamics in human speech motor cortex for a speech prosthesis

项目概要 适用于严重言语障碍人士的增强和替代通信 (AAC) 技术 和运动障碍（SSMI）持续改善，最近在 通信设备的神经控制。在之前 NIDCD 支持的研究中，我们的研究 团队开发了一种高性能皮质内脑机接口（iBCI），可以解码 手臂的运动意图直接来自于大脑的活动。这项技术让人们能够 SSMI 以足够的速度和精度控制计算机光标，最多可输入 8 个字符 字/分钟，并且仅使用解码即可完全控制未修改的消费设备 运动皮质活动。在拟议的 U01 临床研究中，作为多中心研究的一部分进行 BrainGate 联盟，我们将基于数十年研究运动系统的经验 人类和非人类灵长类动物，最终目标是推进 iBCI 技术。的目标 该项目旨在研究如何在合奏层面准备和产生语音 肌萎缩侧索硬化症患者大脑中与语言相关的运动区域的单个神经元 硬化症（ALS），并创造一种能够以一定速度进行交流的言语假体 接近会话式演讲（每分钟 120-150 个单词）。我们将接近这些 使用一套先进方法进行研究，包括（1）新开发的动态方法 系统计算方法为我们的功能提供了基本的见解 运动系统，以及（2）用于解码运动意图的机器学习算法 语言建模已成为迄今为止最快的通信假肢的基础 报道称。最后，我们将继续评估基于 Utah 阵列的 iBCIs 的安全性 通过正在进行的 BrainGate2 试点临床试验。项目完成后将推进 iBCIs 的通信能力和我们对详细神经网络的理解 言语产生的机制。

PROJECT SUMMARY Augmentative and alternative communication (AAC) technology for people with severe speech and motor impairment (SSMI) continues to improve, with recent advances being made in the neural control of communication devices. In prior NIDCD-supported research, our research team developed a high-performance intracortical brain-computer interface (iBCI) that decodes arm movement intentions directly from brain activity. This technology has allowed people with SSMI to control a computer cursor with sufficient speed and accuracy to type at up to 8 words/min and has enabled full control of unmodified consumer devices using only decoded motor cortical activity. In the proposed U01 clinical research, performed as part of the multi-site BrainGate consortium, we will build upon decades of experience in studying the motor system in humans and non-human primates, with the end goal of advancing iBCI technology. The goals of this project are to study how speech is prepared and produced at the level of ensembles of single neurons in speech-related motor areas of the brain in people with amyotrophic lateral sclerosis (ALS), and to create a speech prosthesis that will allow communication at rates approaching conversational speech (120-150 words per minute). We will approach these investigations with a suite of advanced methods, including (1) newly-developed dynamical systems computational approaches that have provided fundamental insights into the function of the motor system, and (2) machine learning algorithms for decoding of movement intention and language modeling that have formed the basis of the fastest communication prosthesis yet reported. Finally, we will continue to evaluate the safety profile of Utah-array based iBCIs through the ongoing BrainGate2 pilot clinical trial. Upon completion, this project will advance both the capabilities of iBCIs for communication and our understanding of the detailed neural mechanisms of speech production.