NCS-FO: Collaborative Research: Micro-scale Real-time Decoding and Closed-loop Modulation of Human Language

人类创造语言，这是我们物种和文明的一个定义性特征。我们可以从一个巨大的词汇库中精确地选择单词，而且错误率非常低。这一复杂的语言产生系统很容易受到疾病的影响，这也许并不奇怪。脑损伤引起的语言障碍影响着数百万美国人，而且几乎没有补救的希望。近年来，对语音产生的解剖学、生理学和计算基础的研究取得了重要进展，但这受到了明显缺乏有关该过程动力学信息的限制。这一限制是由于现有数据收集工具的时空分辨率低以及目前分析工具的有效性。在这个项目中，我们的驱动愿景是在小的时空尺度上对人类语言系统中的皮层连接进行无与伦比的理解。我们拥有丰富的专业知识，在信号解码的过程中的线索词生产与颅内记录技术，以及使用皮层刺激调制系统。FDA批准的阵列将用于在接受颅内电极放置以定位癫痫发作的癫痫患者的语言产生和语言系统的短暂神经调节期间对感觉运动过程进行闭环解码。最终，语言系统中感觉运动回路的细粒度理解和表示需要开发超小型节能检测器，这将使该探索性项目中获得的知识最终应用于持续神经损伤导致普遍语言障碍的患者。这个综合项目将创新的微电子技术与最先进的大数据分析技术结合在一起，开始开发第一个同类系统来治疗语言障碍。工程目标是开发生物相容的微芯片，以大大提高我们对语言和其他认知过程和学习的洞察力。将开发硅技术的微型芯片，可以记录神经信号，对它们进行解码，并将信号无线传输到体外接收器。当PI开发基于微型探测器和神经调制器群体的闭环真实的时间解码和瞬态神经调制系统时，该项目的三重推力将被整合。该系统有可能提供对人类语言系统前所未有的详细了解，并为失语症和其他语言障碍患者的神经修复提供框架和硬件。该项目体现了多个高风险的目标，有可能改变神经工程范式从记录和调制？只有？大脑的几个区域来部署一群超小型和节能的检测器-调制器。

Humans produce language, which is a defining characteristic of our species and our civilization. We can select words precisely out of a large lexicon with remarkably low error rates. It is perhaps not surprising that this complex speech production system is easily affected by disease. Brain damage induced language disorders affect millions of Americans, and there is little hope of remediation. Research on the anatomical, physiological, and computational bases of speech production has made important strides in recent years but this has been limited by a glaring lack of information on the dynamics of the process. This limitation results from the low spatio-temporal resolution of the available tools to collect data and the effectiveness of the current tools for analysis. Our driving vision in this project is to develop an unparalleled understanding of cortical connectivity in the human language system at small spatio-temporal scales. We possess much expertise in signal decoding of the processes of cued word production with intracranial recording techniques, as well as using cortical stimulation to modulate the system. FDA-approved arrays will be used to perform closed-loop decoding of sensorimotor processes during speech production and transient neuromodulation of the language system in patients with epilepsy undergoing intracranial electrode placement for the localization of seizures. Ultimately though, the fine-grained understanding and representation of sensorimotor loops in the language system necessitates the development of ultra-small energy efficient detectors that will enable the knowledge gained in this exploratory project to be eventually applied in patients who have sustained neurological injuries that have resulted in pervasive language impairments. This integrative project brings innovative microelectronics technologies together with state of the art large data analysis techniques to begin to develop a first of its kind system to remediate language disorders.The engineering objective is to develop biocompatible microchips to vastly enhance our insight into language and other cognitive processes and learning. Miniaturized microchips in silicon technology will be developed that can record neural signals, digitize them, and transmit the signals to an in vitro receiver wirelessly. The three-fold thrust of the project will be integrated when the PIs develop closed-loop real time decoding and transient neuromodulation system based on a population of miniaturized detectors and neuromodulators. The system has the potential to provide an unprecedented detailed understanding of the human language system and provide the framework and hardware for neural prosthetics in patients with aphasia and other language disorders. The project embodies multiple high-risk goals that have the potential to shift neuroengineering paradigm from recording and modulating in ?only? a few regions of the brain to deploying a population of ultra-small and energy efficient detectors-modulators.