A Hybrid Neural-Machine Interface for Volitional Control of a Powered Lower Limb Prosthesis

用于动力下肢假肢意志控制的混合神经机器接口

• 批准号: 10214202
• 负责人: Justin Alexander Brantley
• 金额: $ 7.26万
• 依托单位: UNIVERSITY OF PENNSYLVANIA
• 依托单位国家: 美国
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-09-28 至 2023-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10214202
• 关键词: Activities of Daily Living; Affect; Amputees; Ankle; Artificial Intelligence; Award; BRAIN initiative; Bone Diseases; Brain; Brain imaging; Chronic; Communication; Complex; Detection; Development; Electroencephalography; Electromyography; Electronic Mail; Energy Metabolism; Engineering; Frequencies; Funding Agency; Future Teacher; Gait; Gait abnormality; Goals; Grant; Health; Human; Hybrids; Image; Imaging Techniques; Imaging technology; Impairment; Individual; Injury; Intelligence; International; Intuition; Joints; Knee; Knowledge; Leadership; Limb Prosthesis; Link; Locomotion; Lower Extremity; Machine Learning; Measures; Medical; Medical Technology; Mentors; Mentorship; Metabolic; Methods; Mission; Motion; Movement; Muscle; Nervous System Trauma; Neurologic; Neurologic Effect; Neurosciences; Oral; Orthotic Devices; Patients; Pattern; Performance; Population; Positioning Attribute; Prosthesis; Public Health; Ramp; Research; Research Project Grants; Residual state; Robotics; Scalp structure; Signal Transduction; Social Network; Source; Surface; System; Teacher Professional Development; Technical Expertise; Technology; Telephone; Time; Training Programs; Translating; United States; United States National Institutes of Health; Upper Extremity; Visual Cortex; Volition; Walking; Work; arthropathies; base; clinically significant; exhaustion; experience; fall risk; improved; innovation; instrument; kinematics; limb amputation; limb movement; multimodality; neural correlate; neurological rehabilitation; neuroregulation; neurotransmission; novel strategies; physically handicapped; post stroke; posters; powered exoskeleton; powered prosthesis; prediction algorithm; programs; rehabilitation paradigm; relating to nervous system; robot control; robot exoskeleton; robot rehabilitation; robotic system; sensor; signal processing; skills; stroke rehabilitation; symposium; synergism; undergraduate student; visual motor

Project Summary The goal of the proposed work is to develop a robust hybrid neural-machine interface (NMI), combining brain and muscle signals, to improve overall control of a lower limb prosthetic device during activities of daily living. Limb amputation affects over 600,000 individuals annually in the US, and is a major cause of physical disability that causes activities of daily living to become difficult or impossible for the amputee. The limitations of current lower-limb prostheses are associated with limited volitional control, reduced mobility, and chronic gait abnormalities, which have been linked to exhaustion from increased energy expenditure, increased risk of falling, and degenerative bone and joint disorders in both the intact and amputated limb. In this study, EMG signals from both residual and intact lower limbs and EEG signals from the cortex are leveraged to decode transitions to and from various modes of locomotion modes in able-bodied individuals and transfemoral amputees, and to provide a global understanding of movement at the cortical, muscular, and kinematic level in amputees. Specifically, time and frequency domain features are leveraged to create a prediction algorithm capable of identifying upcoming terrain transitions in advance. In lower limb amputees, this hybrid NMI paradigm translates to volitional control of a powered lower-limb prosthesis, which allows for seamless transitions between various movement conditions. The high-level of control is expected to result in significant increases in level of activity and overall improvements in gait. Previous studies have demonstrated the feasibility of EEG or EMG based NMIs for orthotic and prosthetic devices; however, no study to date has integrated EEG and EMG in a NMI for powered lower limb prostheses. This study is motivated by the need to explore advanced neural control sources for intuitive control of artificial limbs. This project aligns directly with the Mission & Goals of the NIH, the Brain Initiative, and NIH’s Blueprint Program by expanding fundamental knowledge of neuroscience, human, health and wellness; by utilizing an innovative research strategy; and ultimately returning the knowledge to the public through the development of a highly advanced medical technology. Furthermore, the technology developed through this work has implications beyond the amputee population in the treatment of many neurological conditions and injuries, such as in neurorehabilitation after stroke. The innovation of this project lies in the novel approach of using multimodal neural signals and movement synergies as a framework for interpreting movement of the lower limb. The scientific impact is realized by a greater understanding of the neural correlates of movement after lower-limb amputation. The direct clinical significance for the patient can be measured directly through improved gait performance and walking confidence, leading to increased mobility and a reduced risk of falling, exhaustion, and bone and joint disorders.

项目摘要 拟议工作的目标是开发一种健壮的混合神经-机器接口(NMI)， 结合大脑和肌肉信号，改善对假肢装置的总体控制 日常生活活动。截肢手术在美国每年影响60多万人，是一项主要的 导致截肢者难以或不能进行日常生活活动的身体残疾原因。 目前的下肢假肢的局限性与意志控制有限、活动能力降低、 以及慢性步态异常，这与能量消耗增加导致的疲惫有关， 增加跌倒的风险，以及完整和截肢肢体的退行性骨骼和关节疾病。在这 研究发现，来自残肢和完整肢体的肌电信号和来自大脑皮质的脑电信号被用来 解码健全个体和经股动脉不同运动模式之间的转换 截肢者，并提供对皮质、肌肉和运动学水平上的运动的全面了解 截肢者。具体地说，利用时间域和频域特征来创建预测算法 能够提前识别即将到来的地形转换。在下肢截肢者中，这种混合NMI范例 转化为对电动假肢的意志控制，这允许在 各种运动条件。高水平的控制预计将导致 步态的活跃度和整体改善。以往的研究已经证明了脑电或肌电的可行性。 基于NMI的矫形和假肢设备；然而，到目前为止，还没有研究将EEG和EMG结合在一起 NMI用于电动假肢。这项研究的动机是需要探索先进的神经控制 用于直观控制假肢的源码。 该项目直接与NIH的使命和目标、大脑倡议和NIH的蓝图保持一致 计划通过扩展神经科学、人类、健康和健康的基础知识；通过利用 创新的研究策略；并最终通过制定一个 非常先进的医疗技术。此外，通过这项工作开发的技术具有深远的影响 在许多神经疾病和损伤的治疗中超越截肢者群体，例如在 中风后的神经康复。该项目的创新之处在于采用了多式联运的新颖方法 神经信号和运动协同作用作为解释下肢运动的框架。这个 科学的影响是通过更好地理解下肢运动后的神经关联来实现的。 截肢。通过改善步态可以直接测量患者的直接临床意义 性能和行走信心，从而提高机动性并降低摔倒、筋疲力尽和 骨骼和关节疾病。