Low-burden, high-throughput brain-computer interfaces

低负担、高通量脑机接口

• 批准号: RGPIN-2019-06033
• 负责人: Chau, Tom
• 金额: $ 4.66万
• 依托单位: University of Toronto
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2021
• 资助国家: 加拿大
• 起止时间: 2021-01-01 至 2022-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=738668
• 关键词: Low; burden; throughput; brain; computer

CONTEXT A brain-computer interface (BCI) facilitates direct communication between the human brain and a computer and as such provides an attractive, potentially highly intuitive means of interacting with technology. Indeed, there has been a surge of interest in BCI applications in diverse industry sectors, including: home automation, military communications, gaming, geriatrics, assistive technology, and law enforcement, among others. However, current BCI technology requires arduous machine calibration and user training before a user's brain signals can be accurately recognized. Further, BCI-based communication remains slow compared to human face-to-face interaction. These challenges need to be overcome before the potential of BCIs can be broadly realized in society. OBJECTIVES The focus of my Discovery Grant is to tackle these challenges, specifically to: 1. Reduce the burden of machine calibration and user training, and 2. Increase the rate of BCI-based communication. PROPOSED RESEARCH To reduce the burden of calibration and training, we will investigate several technological approaches to accelerating machine and user learning. One method is to endow the BCI with a mathematical model developed for other users and to determine a way to efficiently adjust that model to a new user, while the user is also learning under the guidance of an expert. To increase the rate of BCI-based communication, we will deploy algorithms that understand the meaning of words spoken to a BCI user and subsequently present the BCI user with only messages that are relevant to the conversation at hand. We will develop a fast and intuitive way for the BCI user to select the desired message, simply by reading or imagining the words. ANTICIPATED OUTCOMES The proposed research will generate new ways for machines to learn to recognize changing brain patterns of a human user and new algorithms that help human users acquire the necessary skill to interact with technology via mental activity alone. This research will also yield a new and more efficient way to communicate with another human being via a BCI. Collectively, these outcomes will improve the practicality and usefulness of BCI technology. The proposed program will train 8 PhD students and up to 3 post-doctoral fellows. IMPORTANCE This research is of importance to Canadians living with severe disabilities. An easy-to-learn and fast BCI will open doors to educational attainment and eventual employment to those who are unable to communicate through speech or movements. This research will introduce to the BCI research field new theory and algorithms for concurrent machine and user learning, the notion of a BCI coach, evidence about the value of multisensory feedback and potentially new ways to detect user preference through brain signals. Finally, the program will assert a Canadian role in the burgeoning field of BCI by producing technologies that can be transferred to companies and BCI experts to lead the emerging sector.

脑机接口（BCI）促进了人脑和计算机之间的直接通信，因此提供了一种有吸引力的、潜在的高度直观的技术交互方式。事实上，在不同的行业领域，BCI应用的兴趣激增，包括：家庭自动化，军事通信，游戏，老年医学，辅助技术和执法等。然而，目前的BCI技术需要艰苦的机器校准和用户培训，才能准确识别用户的大脑信号。此外，与人类面对面的交互相比，基于BCI的通信仍然很慢。这些挑战需要克服之前，BCI的潜力可以在社会上广泛实现。我的发现补助金的重点是解决这些挑战，具体来说：1。减少机器校准和用户培训的负担，以及2.提高基于BCI的通信速率。为了减轻校准和训练的负担，我们将研究几种加速机器和用户学习的技术方法。一种方法是赋予BCI为其他用户开发的数学模型，并确定一种方法来有效地调整该模型以适应新用户，同时用户也在专家的指导下学习。为了提高基于BCI的通信速率，我们将部署算法，这些算法可以理解对BCI用户所说的话的含义，然后仅向BCI用户呈现与手头对话相关的消息。我们将为BCI用户开发一种快速直观的方法来选择所需的消息，只需通过阅读或想象的话。预期结果拟议的研究将产生新的方法，让机器学习识别人类用户不断变化的大脑模式，以及新的算法，帮助人类用户获得必要的技能，仅通过心理活动与技术进行交互。这项研究还将产生一种新的、更有效的方式，通过BCI与另一个人进行交流。总的来说，这些成果将提高BCI技术的实用性和有用性。该项目将培养8名博士生和3名博士后研究员。这项研究对患有严重残疾的加拿大人很重要。一个易于学习和快速的BCI将为那些无法通过语言或动作进行交流的人打开教育和最终就业的大门。这项研究将向BCI研究领域引入并行机器和用户学习的新理论和算法，BCI教练的概念，多感官反馈价值的证据以及通过大脑信号检测用户偏好的潜在新方法。最后，该计划将通过生产可以转让给公司和BCI专家的技术，在新兴的BCI领域发挥加拿大的作用。