The purpose of the study is to explore the feasibility of using a wearable device, called a myoelectric-computer interface (MCI), to improve arm movement in people who have had a stroke.

Impaired arm movement after stroke is caused not just by weakness, but also by impaired coordination between joints due to abnormal co-activation of muscles. These abnormal co-activation patterns are thought to be due to abnormal movement planning.The MCI aims to reduce abnormal co-activation by providing feedback about individual muscle activations.

This randomized, controlled, blinded study will test the home use of an MCI in chronic and acute stroke survivors.

Inclusion Criteria:

Chronic stroke participants

• Hemiparesis from first ever stroke at least 6 months prior to screening
• Severe motor impairment (FMA of 7-30)
• At least some voluntary shoulder and elbow muscle activation.

Acute stroke participants

• Hemiparesis from first ever stroke within the past 21 days
• Severe motor impairment (FMA of 3-20), or total Manual Motor Score of 1-8 combined in Shoulder Abduction and Finger Extensors

Exclusion Criteria:

• Cognitive impairment with at least moderately impaired attention, or unable to follow instructions of the MCI task
• Visual impairment (such as hemianopia) preventing full view of the screen
• Anesthesia or neglect in the affected arm, or visual hemineglect (score of 2 on the NIH Stroke Scale Extinction and Inattention subtest).
• Participation in another study on the affected arm within 6 weeks of enrollment or any pharmacological study
• Inability to understand or follow commands in English due to aphasia or other reason
• Diffuse or multifocal infarcts
• Substantial arm pain preventing participation for 90 minutes a day
• New spasticity treatment (pharmacological or Botox)

Participants will be blinded as to sham vs. intervention, but not as to which interventional arm.