Upper limb control of robotic lower limb assistance during walking

描述（由申请人提供）：机器人设备在理解神经完整和神经受损个体的运动控制方面取得了重大进展。许多研究使用机器人设备来研究上肢运动的适应和控制，但研究下肢的机器人设备相对较少。Ferris和他的同事开发了机器人下肢矫形器来研究人类运动中的运动适应。初步研究表明，在机器人矫形器的辅助下，人类可以充分适应下肢肌肉激活模式的时间和大小，以接近正常的运动学方式行走。使用上肢控制下肢机器人矫形器的能力可以极大地促进神经损伤患者的运动康复。这种新颖的控制方法可以让患者直接控制机器人辅助的时间和大小。然而，尚不清楚人类是否能够轻易地适应上肢肌肉活动，以有效地控制行走过程中的下肢机器人辅助。因此，拟议研究的总体目标是确定1)人类是否会在跑步机行走时改变上肢肌肉活动和/或运动，以控制机器人踝足矫形器（AFO）的足底屈曲辅助，以及2)一个控制器是否允许用户更快地适应或从AFO获得比其他控制器更多的帮助。我们将让健康的人类受试者穿着由上肢控制的人造足底屈肌驱动的机器人AFO行走。我们将测试机器人AFO的3种不同的比例控制方法：使用肱三头肌的肌电控制，使用肘部伸展的运动学控制和手持按钮。我们的主要变量是AFO在推离过程中所做的净机械功。这将使我们能够量化AFO在行走过程中对受试者的辅助程度。我们还将测试控制器之间在电机适应速度方面的差异。我们将根据达到稳定状态所需的时间来计算运动适应期：a)踝关节运动学相关系数，b)矫形器正功和c)矫形器负功。为了进一步描述受试者如何改变他们的肌肉激活和/或运动，我们还将分析下肢肌肉活动、运动学和动力学。这些研究的结果可能会导致更好的机器人界面，以协助脊髓损伤，中风或其他神经损伤后的步态康复。

DESCRIPTION (provided by applicant): Robotic devices have led to major advances in understanding the motor control of neurologically intact and neurologically impaired individuals. While many studies have used robotic devices to study upper limb motor adaptation and control, there are relatively few robotic devices for studying the lower limb. Ferris and colleagues have developed robotic lower limb orthosis to study motor adaptation during human locomotion. Initial studies indicate that humans can substantially adapt the timing and magnitude of lower limb muscle activation patterns to walk with near normal kinematics while being assisted by the robotic orthosis. The ability to use upper limbs to control lower limb robotic orthosis could greatly facilitate locomotor rehabilitation in individuals with neurological impairment. This novel control method would give the patient direct control over timing and magnitude of the robotic assistance. However, it is not known if humans can readily adapt upper limb muscle activity to effectively control lower limb robotic assistance during walking. Therefore, the overall objectives of the proposed research are to determine 1) if humans will modify upper limb muscle activity and/or movement during treadmill walking to control plantar flexion assistance from a robotic ankle-foot orthosis (AFO), and 2) if one controller allows the user to adapt faster or gain more assistance from the AFO than the other controllers. We will have healthy human subjects walk while wearing a robotic AFO powered by an artificial plantar flexor muscle controlled by the upper limb. We will test 3 different proportional control methods for the robotic AFO: myoelectric control using the triceps muscle, kinematic control using elbow extension, and a handheld pushbutton. Our primary variable will be net mechanical work done by the AFO during push-off. This will allow us to quantify how well the AFO can assist the subject during walking. We will also test for differences between controllers in terms of speed of motor adaptation. We will calculate motor adaptation period by the time required to reach steady state in: a) ankle kinematic correlation coefficient, b) orthosis positive work and c) orthosis negative work. To further characterize how subjects modify their muscle activation and/or movement, we will also analyze lower limb muscle activity, kinematics, and kinetics. The results of these studies could lead to better robotic interfaces for assisting gait rehabilitation after spinal cord injury, stroke, or other neurological impairments. Lay summary: The insight into the neural control of human locomotion and motor adaptation provided by this study may translate into innovative rehabilitation therapies and major advances in powered orthoses and prostheses for individuals with impaired mobility due to neurological injury or amputation.