The aging veteran population is at increased risk for stroke. Each year 400,000 people in the United States survive strokes and the number of stroke survivors now approaches three million. Although most survivors regain the ability to walk, 30% to 60% lose the use of the hemiparetic arm. The common practice of teaching compensatory techniques with the stronger arm is being challenged by studies showing forced use of the hemiparetic limb leads to enhanced cortical reorganization and function. Such therapy is time-consuming, costly, and often limited in duration by financial resources. Semi-automated therapy devices have the potential to increase the amount of therapy received and improve recovery without placing a large burden on the clinical staff. Building on the expertise and experience of Center investigators and Stanford collaborators in assistive robotics, we established a research program to develop and evaluate robots for rehabilitative therapy. Robotic manipulators offer 3 significant advantages over previous methods used to study and treat hemiparesis: 1) interaction force and torque measurements serve as sensitive, continuous measures of motor recovery, 2) well-controlled, repetitive sensorimotor stimulation can be delivered at high intensity and at reasonable cost, and 3) novel treatment protocols can be readily designed and objectively tested. Investigators at Burke Rehabilitation Hospital have demonstrated that adding robot-assisted, planar, unilateral tracking movement to upper limb rehabilitation therapy 2 to 3 weeks post-stroke improves motor recovery. Our previous work with a bimanual, robotic device limited to planar movements (Limb Manipulation under Patient Control: A Pilot Study; Co-PI: Charles G. Burgar, MD and H.F. Machiel Van der Loos, PhD; VA RR&D Pilot project B1846PA) demonstrated that patient-guided manipulation therapy of the paretic limb is feasible, and that interaction forces between the robot and subject's paretic arm correlate with the state of motor recovery, as measured by the upper limb portion of the Fugl-Meyer scale (F-M). With continued VA support (Mechanically-assisted Upper Limb Manipulation for Assessment and Therapy; Co-PI: Charles G. Burgar, MD and H.F. Machiel Van der Loos, PhD; VA RR&D Merit Review project B2056-RA), we developed an advanced robotic system (MIME) that assists or resists elbow and shoulder movements in three-dimensional space. In addition to 3 unilateral exercise modes, a novel bimanual mode enables hemiparetic subjects to practice mirror-image upper limb exercises. In a prospective, blinded, controlled clinical trial, we compared robot-assisted therapy to conventional NDT-based therapy of the same intensity and duration. While both groups had significantly improved F-M scores, the robot-assisted group showed a trend for greater overall improvement than the control group. For the shoulder/elbow portion of the F-M, gains in the robot group were significantly higher than the control group. Post-hoc analysis suggests possible mechanisms for the improvements, including strength gains and improved motor control in paretic muscles, with increased agonist and decreased antagonist activity. This program thus accomplished a major milestone by demonstrating, for the first time, that robot-assisted neuro-rehabilitation not only produces significant improvements in motor recovery in persons for whom spontaneous neurological recovery has reached a plateau, but that therapy with MIME is even more effective than the conventional treatment. Plans are in place to replicate MIME for clinical trials at two other VA facilities. The PIs are also participating in a national consortium planning a multi-center clinical trial of robot-assisted neuro-rehabilitation for acute stroke patients. |
