Adaptive ankle robot control system to reduce foot-drop in chronic stroke

 描述(由申请人提供)： 这项建议研究了一种新的踝关节机器人(Anklebot)自适应控制方法，将其与跑步机训练相结合，以减少脚下垂并改善慢性偏瘫中风幸存者的活动功能。目前，有足部下垂的中风幸存者接受了使用拐杖或其他辅助设备的训练，为了安全起见，他们经常接受脚踝矫形器(AFO)的训练。两者都不能调节任务练习或神经运动恢复。我们已经开发了一种自适应Anklebot控制器，它可以检测步态循环子事件，用于分级机器人辅助的精确计时，从而能够在行走环境中调整缺陷严重程度的脚踝运动学习。我们的优点试点研究结果显示，6周的带机器人的跑步机训练(TMR)仅用于辅助摆动阶段的背屈，比单独使用TM更有效地改善自由行走时的摆动背屈(+400%，+9.7°，N=4/组)和落地行走速度(21%对9%)，并在训练后6周保持优势，4名参与者中有3名不再需要AFO。值得注意的是，摆动阶段的TMR训练改善了偏瘫腿的推离(277%对2%)。压力中心(COP)对站立平衡的影响(-25%比+30%)，这表明除了机器人针对脚部下垂的因素外，其他步态和平衡元素也有普遍好处。这项随机研究调查了这样一种假设，即与单独使用TM相比，6周TMR更有效地持续改善有背屈功能障碍的慢性中风幸存者的步态生物力学、静态和动态平衡以及活动功能。目的是确定6周TMR与单独TMR在以下方面的比较效果：1)以步态速度、摆动阶段DF、终端姿势推开和48小时自由生活活动度曲线为指标的独立步态功能。2)平衡功能，以姿势摆动(COP)、安静站立时的不对称负荷、非瘫痪步态开始时的最大偏瘫A-P力和标准化的量表(Berg平衡、动态步态指数和ABC Falls疗效和ABC)为指标；3)长期活动结果，通过在正式训练停止后6周和6个月重复测量所有关键步态和平衡结果来评估。尽管上肢机器人已被证明是有效的，并改变了中风的国家护理标准，但相比之下，主要专注于重复的、多关节的步态周期模式的下肢机器人仍然存在争议，人们达成了共识 目前的方法不如通常的护理，甚至是有害的。我们通过测试一个自适应控制的Anklebot来挑战这一范式，该机器人由感觉运动学习模型指导，专门专注于由于偏瘫DF控制受损而引起的并发症。基于重复测量的无辅助行走的踝关节运动学习的轮廓显示出一个幂规律模式，大约80%的增长和3周。因此，TMR在常规物理治疗的时间框架内可能被证明是有效的，这增加了转化为VA护理的可能性。这项研究的结果将为不同的缺陷严重程度定制步态集成的自适应模块化LE机器人建立一个新的范例，以改善中风和其他神经疾病的运动功能。

 DESCRIPTION (provided by applicant): This proposal investigates a novel ankle robot (anklebot) adaptive control approach integrated with treadmill training to reduce foot drop and improve mobility function in chronic hemiparetic stroke survivors. Currently, stroke survivors with foot drop are trained to live with a cane or othr assistive device, and often ankle foot orthotics (AFO's) for safety. Neither mediates task-practice or neuromotor recovery. We have developed an adaptive anklebot controller that detects gait cycle sub-events for precise timing of graded robotics assist to enable deficit severity adjusted ankle motor learning in the context of walking. Our Merit pilot findings show that 6 weeks treadmill training with robot (TMR) timed to assist swing phase dorsiflexion only, is more effective than TM alone to improve free-walking swing dorsiflexion at foot strike (+400%, +9.7°, N=4/group), floor-walking speed (21% vs. 9%), and the benefits are retained at 6 weeks post- training such that 3 of 4 participants no longer required their AFO's. Notably, swing-phase TMR training improved paretic leg push-off (277% vs. 2%), and center-of-pressure (CoP) sway on standing balance (-25% vs. +30%), indicating generalized benefits to other elements of gait and balance, beyond those robotically targeted toward foot drop. This randomized study investigates the hypothesis that 6 weeks TMR is more effective to durably improve gait biomechanics, static and dynamic balance, and mobility function in chronic stroke survivors with dorsiflexion deficits, compared to TM alone. Aims are to determine the compare effectiveness of 6 weeks TMR vs. TM alone on: 1) Independent gait function indexed by gait velocity, swing-phase DF, terminal stance push off, and 48- hour free-living mobility profiles. 2) Balance function indexed by measures of postural sway (CoP), asymmetric loading in quiet standing, peak paretic A-P forces in non-paretic gait initiation, and standardized scales (Berg Balance, Dynamic Gait Index, and ABC for Falls Efficacy and; 3) Long-term mobility outcomes, assessed by repeated measures of all key gait and balance outcomes at 6 weeks and 6 months after formal training cessation. While upper extremity robotics has proven effective, and altered national care standards for stroke, in contrast, lower extremity robotics that has primaril focused on repetitive, multi-joint patterning of gait cycles, remains controversial, with consensus that current approaches are inferior to usual care, or even deleterious. We challenge this paradigm by testing an adaptively controlled anklebot guided by sensorimotor learning models to focus specifically on complications due to impaired paretic DF control. The profile of ankle motor learning based on repeat measures of unassisted walking reveals a power law pattern with ~80% of gains <3 weeks. Hence, TMR may prove effective within the time-frame of usual physical therapy, which increases potential for translation into VA care. Results of this study wil establish a new paradigm for diverse deficit severity customized gait-integrated adaptive modular LE robotics to improve mobility function in stroke and other neurological conditions.