Ankle Robotics After Stroke

This proposal investigates a portable ankle robot (AMBLE) to be used during over-ground mobility training to reduce foot drop and improve walking function in hemiparetic (half-body, partially paralyzed stroke patients with foot drop (inability to properly lift and clear the foot during walking. About 30% of stroke survivors are left with permanent ankle weakness that impairs their mobility and increases fall-risk. Currently, stroke survivors with foot drop live with a cane or other assistive device, and often ankle-foot braces (AFOs) for safety. These assistive devices do not reverse or reduce the underlying neurological foot drop problem. Recognizing the crucial role of ankle function in walking and balance, and recognizing that the distal part of the lower extremity often suffers the greatest damage after a human stroke, the investigators have come up with a portable ankle robot as a tool for therapists to help shape recovery of walking.

The AMBLE, and its underlying control system, uses information about how patients are walking from one step to another to assist and shape foot lifting so as to help re- train walking recovery by a process that neuroscientists call motor learning. It is the combination of the partially paralyzed stroke survivor's movement efforts with timely assistance "only as needed" by the robot that investigators and others show is the key to movement recovery after stroke. Thus, the ankle robot is not a crutch, but a learning and measuring device that incrementally "gets out of the way" of the learner to facilitate human robot learning such that the human takes over more of the volitional learning.

The research team at University of Maryland has demonstrated in 4 prior studies using seated and treadmill based robot assisted training using a bulky laboratory robot programmed with a motor learning formula that can improve ankle motor control in both the early and chronic phases of stroke, and this can improve over-ground unassisted walking. A significant proportion of stroke survivors showed session by session recovery of volitional (not assisted by the robot) ankle lifting during walking across 6 weeks of three 30- 45 minute sessions of robot training while walking on a treadmill, even years after their stroke. In fact, it has been found that two weeks of 3 sessions per week ankle robotics training was the time profile for most motor learning recovery to reduce foot drop. This information has informed the design of the study described below.

Previous research was done using a bulky, heavy (~8 lbs), and expensive laboratory robot that only allowed seated or treadmill based training because it was tethered by wires. This greatly limits how it can be used by physical therapists, and is not appropriately configured for ease of use by physical therapists in practice. NextStep Robotics invented and built the ankle robots motor learning programs with a lot of input from physical therapists and other rehabilitation clinicians into a portable lightweight robot that can be used over-ground anywhere with blue tooth controls that also tell the therapist precisely how well the stroke survivors is learning, step by step. It is this new portable ankle robot that is configured for use in practice that investigators seek to test in studies with physical therapists using it fully integrated into their usual outpatient stroke mobility recovery training at University of Maryland Orthopedics and Rehabilitation Institute.

This U44 Award from the National Institute of Neurological Disorders and Stroke (NINDS) is not a typical single phase randomized clinical study, but consists of Phase I that completes commercial design of the robot the first year, followed by Phase II randomized clinical trial across years 2-4 of a finalized commercial version of the ankle robot.

Phase II (following completion of commercial design in Phase I) is a randomized (group assignment by chance), blinded (outcome testing done by technicians unaware of patient group assignment), two arm (2 groups) study that investigates the hypothesis that in subacute (6 weeks to 6 months) stroke subjects with foot drop, AMBLE integrated physical therapy (PTR) consisting of 18 training sessions over nine weeks is more effective than usual physical therapy (PT) to improve foot drop outcomes measured by movement analyses of walking, and by standardized clinical assessments of walking including specific foot drop outcomes as assessed by a certified PT clinician (blinded to treatment assignment), cross checked by blinded review of 2 other clinicians of films of the standardized mobility assessments to provide a consensus impartial judgement. Notably, this Phase II study focuses on sub-acute stroke recovery (6 weeks to 6 months) because it represents a therapeutic window into which conventional outpatient physical therapy is typically front-loaded to optimize outcomes. This phase of stroke rehabilitation is selected to test the AMBLE in real world settings when outpatient physical therapy typically occurs, using a treatment frequency and duration (18 physical therapy sessions across 9 weeks) that is representative of practice in Maryland and most of the United States. If wearing the AMBLE robot during physical therapy in this time-frame reduces foot drop and improves longer term outcomes measured 3 months after all robotics therapy has ended, then the investigators will apply for FDA approval for the first robotics device to actively treat foot drop after stroke.