Mobile Signal Processing System for Broadband Neural Decoding

描述(由申请人提供)： 开发技术以帮助严重行动障碍、瘫痪、闭锁综合症或截肢的人是退伍军人事务部的优先事项。肌萎缩侧索硬化症(Lou Gehrig‘s病)会导致完全瘫痪，在退伍军人中的发病率高得不成比例。由于在最近的冲突中受伤，对上肢截肢退伍军人的护理正在增加。退伍军人管理局拥有先进的辅助技术，包括ALS患者通信软件的直接神经控制、革命性假肢和瘫痪肢体的功能性电刺激(FES)。BrainGate2(IDE)是一种可植入的大脑信号传感器，与神经解码系统相结合，正在进行的试点临床试验的证据表明，瘫痪患者可以在脊髓损伤、脑干中风或ALS发作数年后使用大脑信号控制软件或机器人/假肢手臂。尽管直接大脑控制承诺对使能技术进行有效、熟练的控制，但目前的神经解码算法运行在笨重、固定的计算机系统上。将这些平台减少到一个紧凑的可穿戴设备将实现里程碑式的进步，可以使瘫痪患者能够在大脑控制的情况下在轮椅上独立移动，允许闭锁综合征患者在任何地方使用神经控制的通信工具，允许活动的肢体截肢患者用自己的思想控制先进的假肢，如DEKA手臂/手，或者让上肢瘫痪患者使用大脑控制的FES用自己的肌肉伸手和抓握。这项拟议的研究通过利用商用移动微系统技术来生产一种电池供电的高性能可穿戴设备来实现这一目标，该设备能够无线接收、处理和解码大脑信号，并生成命令来操作附近的移动辅助技术。首先，强大的新一代可编程门阵列(FGA)允许将目前运行在Windows计算机上的神经解码算法软件转换为硬件描述语言(HDL语言)，这种语言在指甲大小的FPGA芯片上运行速度快几个数量级。为了进行研究，可以很容易地对现场可编程门阵列重新编程，以测试新的神经解码算法。其次，为移动应用开发的高性能、低功耗处理器提供了接收神经信号和传输命令所需的数据管理和无线通信能力。集成系统将在可穿戴的电池供电组件中执行目前的BrainGate2神经接口系统所需的所有信号处理和解码功能。将开发接口软件，允许工程师快速重新配置设备，并允许残疾人(或他们的助手)在使用期间调整设备。功能将在BrainGate模拟环境中进行验证。然后，在BrainGate2试验中，四肢瘫痪患者将在控制辅助软件和假肢设备的情况下进行测试。研究将在普罗维登斯退伍军人医学中心和布朗大学进行。这个研究团队在开发创新的微电子和神经假体系统方面拥有扎实的专业知识。这位首席研究员多年来一直指导BrainGate系统工程和开发，在神经假体研究和工业开发创新微电子系统方面都具有生产力。PI和合作研究人员展示了将工程学、神经科学、计算机科学和临床专业知识整合在一起的始终如一的能力，以提供有意义的前沿研究，正在转变神经假体技术，以帮助严重运动障碍的人。通过整合这些专业知识，目前的提议将产生一种可穿戴设备，具有实时解码神经信号所需的前所未有的数据吞吐量和处理性能，使残疾用户能够控制启用的移动辅助设备。

DESCRIPTION (provided by applicant): Developing technologies to assist persons with severe movement disabilities, paralysis, locked-in syndrome or limb amputations is a priority for the Department of Veterans Affairs. ALS (Lou Gehrig's disease), which results in complete paralysis, has a disproportionately high incidence among the veteran population. Due to injuries sustained in recent conflicts, care for veterans with upper extremity amputation is increasing. The VA has advanced assistive technologies including direct neural control of communication software for persons with ALS, revolutionary prosthetic arms, and functional electrical stimulation (FES) of paralyzed limbs. Evidence from the ongoing pilot clinical trial of BrainGate2 (IDE), an implantable brain signal sensor coupled with a neural decoding system, indicates that individuals with paralysis can use their brain signals to control software or robotic/prosthetic arms years after spinal cord injury, brain stem stroke, or the onset of ALS. Although direct brain control promises effective, adept control of enabling technologies, current neural decoding algorithms run on cumbersome, immobile computer systems. Reducing these platforms to a compact, wearable device would achieve cornerstone advances that could enable persons with paralysis to move themselves in a wheelchair independently under brain control, allow persons with locked- in syndrome to access neurally-controlled communication tools anywhere, permit ambulatory individuals with limb amputation to control advanced prostheses such as the DEKA arm/hand with their own thoughts, or let people with upper limb paralysis reach and grasp with their own muscles using brain-controlled FES. The proposed research achieves this goal by exploiting commercially-available mobile microsystem technology to produce a battery-powered, high-performance wearable device capable of wirelessly receiving, processing, and decoding brain signals and generating commands to operate nearby mobile assistive technologies. First, a powerful new generation of programmable gate arrays (FPGAs) allows neural decoding algorithm software that currently runs on a Windows computer to be converted to a hardware description language (HDL) that runs orders of magnitude faster on a fingernail-sized FPGA chip. For research, FPGAs can readily be re-programmed to test novel neural decoding algorithms. Second, high-performance, low- power processors developed for mobile applications provide the requisite data management and wireless communication capabilities to receive neural signals and transmit commands. The integrated system will execute all signal processing and decoding functions required by the present BrainGate2 neural interface system yet in a wearable, battery powered package. Interface software will be developed to allow engineers to rapidly reconfigure the device and to allow individuals with disability (or their assistants) to adjust the device durig use. Functionality will be validated in the BrainGate simulation environment. Then, individuals with tetraplegia in the BrainGate2 trial will test it while controlling assistive software and prosthetic devices. Research will be performed at the Providence VA Medical Center and Brown University. This research team has well-establish expertise in the development of innovative microelectronics and neural prosthetic systems. The Principle Investigator has directed BrainGate systems engineering and development for years with productivity both in neural prosthetics research and in industry developing innovative microelectronic systems. The PI and co-investigators have demonstrated the consistent ability to integrate engineering, neuroscience, computer science, and clinical expertise to deliver meaningful leading-edge research that is transforming neural prosthetic technology to assist persons with severe movement disability. By integrating this expertise, the current proposal will yield a wearable device with the unprecedented data throughput and processing performance required for real-time decoding of neural signals to enable users with disability to control enabling, mobile assistive devices.