NSF-FR: Bidirectional Neural-Machine Interface for Closed-Loop Control of Prostheses

NSF-FR：用于假肢闭环控制的双向神经机器接口

• 批准号: 2319139
• 负责人: Xiaogang Hu
• 金额: $ 399.96万
• 依托单位: Pennsylvania State Univ University Park
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-09-15 至 2028-08-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2319139&HistoricalAwards=false
• 关键词: NSF; FR; Bidirectional; Neural; Machine

Humans can control their limbs to perform a variety of daily tasks with great precision and remarkable adaptability in unpredictable environments, thanks to our cognitive capacity and physical characteristics. People with disabilities could rely on assistive robots that have similar functional capabilities of real limbs, yet people find it difficult to use on a daily basis, partly because the interfaces are unnatural and unintuitive. The objective of this project is to understand the neural and cognitive processes brought to bear during daily tasks, such as reaching and grasping, and to establish natural and nature-inspired approaches that allow the user and the machine (the artificial limb) to communicate. The research outcomes will reduce motor disability and improve quality of life of individuals with physical disabilities. The developed approaches can also enable intuitive control of assistive robots in medical, industrial, and military applications. Summer projects and outreach events, incorporating the proposed techniques, will be offered to undergraduate students in minority-serving universities and local K-12 students, specifically targeting underrepresented students. The research team will organize workshops at national conferences to disseminate research findings and facilitate broader collaborations. Certificate and credential programs will be offered through online learning platforms. Research outcomes will also be presented to local and regional patient support groups and national clinical-oriented conferences so as to disseminate state-of-the-art research development to end users.The goal of this project is to develop and evaluate a biomimetic human-centric neural-machine interface system, which incorporates outward (efferent) and inward (afferent) directed signals for the control of assistive robots. The system will allow individuals with disabilities to interact with their assistive robots as they use their biological limbs. If successful, it will provide a robust and effective model for intuitive interaction of human-machine systems for application to a broader variety of health and industrial applications, and finally overcome the problem of intuitive control of assistive devices in individuals with disability. The research team will strategically integrate research threads that address critical barriers for human-robot integration: Thread 1 will develop implantable and wearable electrode platforms for neural recording and neural stimulation. Thread 2 will understand fundamental principles of neural encoding of artificial sensation and establish biomimetic sensory encoding strategies. Thread 3 will develop an integrated shared control framework for dexterous control of robotic hands. Thread 4 will collectively address the functional integration of closed-loop robotic systems for perceptual motor control. The research team will integrate the proposed techniques, closing the loop between artificial sensing and actuation of the robot and the perception and control authority of the human, examining the adaptability and robustness of the closed-loop human-machine systems. Collectively, the research project can generate transformative outcomes that can blur the boundary between humans and assistive robots, allow end-users to fully leverage the functionality of advanced robots, and promote the development of next-generation neural-machine interfaces and assistive robots.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.

凭借我们的认知能力和身体特征，人类可以在不可预测的环境中以极高的精度和卓越的适应性控制自己的四肢执行各种日常任务。残疾人可以依赖具有与真实肢体类似功能的辅助机器人，但人们发现日常使用起来很困难，部分原因是界面不自然且不直观。该项目的目标是了解日常任务（例如伸手和抓握）中所涉及的神经和认知过程，并建立自然和受自然启发的方法，使用户和机器（假肢）能够进行交流。研究成果将减少运动障碍并提高身体残疾者的生活质量。所开发的方法还可以实现医疗、工业和军事应用中辅助机器人的直观控制。将向少数族裔大学的本科生和当地 K-12 学生提供纳入拟议技术的夏季项目和外展活动，特别针对代表性不足的学生。研究团队将在国家会议上组织研讨会，以传播研究成果并促进更广泛的合作。证书和证书课程将通过在线学习平台提供。研究成果还将提交给当地和区域患者支持小组和国家临床会议，以便向最终用户传播最先进的研究进展。该项目的目标是开发和评估仿生人体中心神经机器接口系统，它结合了向外（传出）和向内（传入）定向信号来控制辅助机器人。该系统将允许残疾人在使用其生物肢体时与辅助机器人互动。如果成功，它将为人机系统的直观交互提供一个强大而有效的模型，以应用于更广泛的健康和工业应用，并最终克服残疾人辅助设备的直观控制问题。研究团队将战略性地整合研究线索，以解决人机集成的关键障碍：线索 1 将开发用于神经记录和神经刺激的可植入和可穿戴电极平台。主题 2 将了解人工感觉神经编码的基本原理并建立仿生感觉编码策略。 Thread 3将开发一个集成的共享控制框架，用于灵巧地控制机器人手。 Thread 4 将共同解决用于感知电机控制的闭环机器人系统的功能集成。研究团队将整合所提出的技术，闭合机器人的人工感知和驱动与人类的感知和控制权限之间的循环，检验闭环人机系统的适应性和鲁棒性。总的来说，该研究项目可以产生变革性成果，模糊人类和辅助机器人之间的界限，让最终用户充分利用先进机器人的功能，并促进下一代神经机器接口和辅助机器人的发展。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。