Of the 5.7 million stroke survivors in the United States, up to 80% exhibit significant weakness in one arm (called "hemiparesis"). This devastating impairment undermines performance of valued activities and quality of life. Although rehabilitation is commonly provided, conventional affected arm rehabilitative strategies have negative evidence, or no evidence, supporting their use. Thus, there remains a need for evidence-based rehabilitative strategies for arm hemiparesis.

Newer rehabilitative approaches emphasize repetitive, task-specific practice (RTP) incorporating the affected arm. However, many of these promising regimens require participation in intensive therapies, and most are only efficacious on the least impaired patients. Thus, there remains a need for an efficacious, practical RTP technique to address moderate affected arm hemiparesis.

To address the above shortfalls, one of the investigators team members piloted an innovative brace integrating electromyography (EMG) and robotics. In his case series, 8 stroke patients exhibiting moderate arm impairment successfully participated in RTP, with the brace (called the "Myomo") detecting and augmenting their movement attempts. Aided by the Myomo, participation in the RTP regimen reduced subjects' affected arm impairment and spasticity. The next logical step is to test Myomo + RTP efficacy using randomized controlled methods and an appropriate sample size.

The primary objective of this small efficacy study is to collect randomized, controlled pilot data estimating clinical effectiveness of this combined, EMG/robotics approach in conjunction with RTP. A unique adjunct will be acquisition of biomechanical data as an outcome measure. Based on pilot data, the central hypothesis is that Myomo therapy with RTP will exhibit significant impairment reductions. We will test the central hypothesis and accomplish the objective by pursuing the following specific aims:

Specific Aim 1: Compare efficacy of Myomo + RTP with RTP only and Myomo only on affected arm impairment. Thirty stroke patients exhibiting moderate affected arm deficits will be randomly assigned to receive: (a) Myomo combined with RTP (Myomo + RTP); (b) RTP only (RTP), which constitutes the most frequently used regimen in clinical environments;2,3 or (c) Myomo therapy only (Myomo). The current study design will determine the additive effect of Myomo training to RTP, while also discerning the relative efficacy of RTP only or Myomo use only. The intended duration of contact will be equivalent across groups.

Affected arm impairment, the primary study outcome, will be measured by the upper extremity section of the Fugl-Meyer Impairment Scale (FM), the primary outcome measure. The FM will be administered by a blinded rater twice before intervention, immediately post-intervention (POST), and two months post-intervention (POST-2) with POST serving as the primary study endpoint. The primary study hypothesis is that subjects in the Myomo + RTP group will exhibit significantly larger FM score changes than Myomo or RTP subjects at POST. It is also hypothesized that these changes will be sustained at POST-2.

Specific Aim 2: Determine the impact of Myomo + RTP on affected arm outcomes. The investigators will administer the Arm Motor Ability Test (AMAT), to all subjects before intervention, at POST, and at POST-2. When compared to their scores before intervention, it is hypothesized that Myomo + RTP subjects will exhibit significantly larger AMAT score changes,than subjects in the other groups at POST. These differences will be sustained at POST-2.

Inclusion Criteria:

• upper extremity Fugl Meyer score >10-< 25 (i.e., subject must score between (and including) 10 and 25 on the UE FM at the baseline screening appointment only)
• presence of volitionally activated EMG signal from the paretic biceps brachii of at least 5 ųV in amplitude
• stroke experienced > 1 month prior to study enrollment
• a score > 24 on the Folstein Mini Mental Status Examination (MMSE)
• age > 35 < 85
• have experienced one stroke
• discharged from all forms of physical rehabilitation
• Myomo brace fits on affected arm properly and without discomfort (i.e., no red marks or discomfort observed in 10 minutes of use during fitting).
• < 35 years old
• excessive pain in the affected hand, arm or shoulder, as measured by a score > 5 on a 10-point visual analog scale
• excessive spasticity at the affected elbow, as defined as a score of > 4 on the Modified Ashworth Spasticity Scale
• currently participating in any experimental rehabilitation or drug studies
• apraxia (< 2.5 on the Alexander scale)
• severe sensory loss in affected hand (Nottingham sensory scale at least 75% of normal)
• severe language deficits (score < 2 on NIH Stroke Scale question 9)
• Stroke that occurred in the brainstem (corticospinal tracts are the final pathway for the motor system, and are frequently damaged in brainstem strokes. These individuals are hypothesized to be less likely to benefit from the cortical plasticity seen with exercise therapy, and are therefore excluded from this study.)
• A current medical history of uncontrolled cardiovascular, or pulmonary disease, or other disease that would preclude involvement in a therapeutic treatment (Subjects must be able to tolerate a one-hour upper-extremity therapy session.)
• History of neurological disorder other than stroke (other neurological disorders may affect the upper extremity motor performance of subjects.); (11) Other significant pain or skin irritation in the upper extremity that would be exacerbated if by the use of the brace (While wearing the brace, the system is pressed close to the upper arm and strapped around it. (If the subject suffers from dermal breakdown or other skin conditions that may be aggravated by such a situation, they should not be involved in the study.)
• Substantial contracture of elbow, defined as > 20 degrees of elbow flexion, as measured at the baseline evaluation. (The system cannot work in the presence of reduced range of motion due to contractures.).