Robotic Training for Stroke Neurorehabilitation

The overall goal is to further develop a robotic exoskeleton for neurorehabilitation of arm function after stroke. The investigators previously developed a novel training protocol that combined the ARMin and HandSOME exoskeletons. This is one of only a few arm exoskeletons that allow coordinated whole limb training in reach and grasp tasks with both virtual and real objects. However, the robot has a very large inertia and friction, and only gross grasp patterns are available. In development work, the investigators will significantly modify the AMRin and HandSOME to deliver a state-of-the-art lightweight robotic exoskeleton capable of retraining a wide range of functional activities. In the subsequent testing phase, a clinical trial will examine the effects of robotic training in chronic stroke subjects.

In a pilot clinical trial, the investigators found that the current robotic exoskeleton elicited improvements in arm function that can potentially supplement conventional methods to improve outcomes. However, the robot has a very large inertia and friction, and only gross grasp patterns are available. Therefore, improvements in movement speed and fine grasp were limited after robotic training. In development work, the investigators will significantly modify the ARMin and HandSOME to deliver a state-of-the-art lightweight robotic exoskeleton capable of retraining a wide range of functional activities. The investigators will reduce the inertia and friction of the robot to 1/4 of current values, incorporate an adaptive algorithm to automatically adjust assistance levels and extend the range of grasp patterns to include power grasp, thumb-index finger pinch and key pinch. Using the improved device from the development activities, the investigators will perform a clinical trial to compare the effectiveness of robotic training to conventional therapy from an occupational therapist. To take advantage of the facilitatory effect of robot therapy on subsequent conventional therapy, the experimental treatment will be 12 hours of robot therapy followed by 12 hours of conventional therapy. Chronic stroke subjects (N=38) will be randomly assigned to receive this experimental treatment or 24 hours of conventional therapy from an occupational therapist.

The experimental group will receive 1-hour robotic training sessions, 3 times per week for a total of 12 sessions supervised by a research assistant. Immediately following this robot training, these subjects will receive the same dosage and schedule (1-hour sessions, 3 times/week, 12 total sessions) of conventional one-on-one therapy from an occupational therapist.

Subjects will receive 24 hours of one-on-one treatment from an occupational therapist. The treatment schedule will parallel that given to the experimental group (1-hour sessions, 3 times/week).

Subjects will be placed in the robot and practice common upper extremity tasks involving grasping, manipulating and moving objects.

An occupational therapist will provide one-on-one individualized programs focused on arm function. Treatment will focus on practice of specific tasks, such as reach, grasp, transport and release of various objects between different targets. Progression is done by varying the shape, size and weight of objects, altering the end range of the target or increasing the speed of movement.

The Fugl-Meyer Test of Motor Function will be used to assess motor impairments at the shoulder, elbow, wrist and fingers. The Fugl-Meyer test scores reflexes and the ability to perform several movements and tasks on a 3-point scale. The Fugl-Meyer was designed for the recovery patterns observed after stroke and is very responsive to change in severe and moderately impaired subjects.

The Action Research Arm Test (ARAT) is an impairment level measurement tool that assesses the functional limitations of the upper extremities. It is one of the most frequently used primary endpoints in upper extremity (UE) training trials in stroke. The assessment incorporates 19 items that are divided into four subscales: Grasp, Grip, Pinch, and Gross movement. Item scores are summed to form a subtest score, and then a full-scale score.

The Motor Activity Log (MAL) will be used to assess use of the limb at home. It is a structured interview during which respondents are asked to rate how they use their more-impaired arm for 28 activities of daily living (ADL) in the home over a specified period. Activities include brushing teeth, buttoning a shirt or blouse, and eating with a fork or spoon.

The Fugl-Meyer Test of Motor Function will be used to assess motor impairments at the shoulder, elbow, wrist and fingers. The Fugl-Meyer test scores reflexes and the ability to perform several movements and tasks on a 3-point scale. The Fugl-Meyer was designed for the recovery patterns observed after stroke and is very responsive to change in severe and moderately impaired subjects.

The Action Research Arm Test (ARAT) is an impairment level measurement tool that assesses the functional limitations of the upper extremities. It is one of the most frequently used primary endpoints in upper extremity (UE) training trials in stroke. The assessment incorporates 19 items that are divided into four subscales: Grasp, Grip, Pinch, and Gross movement. Item scores are summed to form a subtest score, and then a full-scale score.

The Motor Activity Log (MAL) will be used to assess use of the limb at home. It is a structured interview during which respondents are asked to rate how they use their more-impaired arm for 28 activities of daily living (ADL) in the home over a specified period. Activities include brushing teeth, buttoning a shirt or blouse, and eating with a fork or spoon.

Inclusion Criteria: Age 21 or older Ischemic or hemorrhagic stroke (with confirmatory neuroimaging) that occurred more than 6 months before entering the study Presence of voluntary hand activity indicated by a score of at least 1 on the finger mass extension/grasp release item of the Fugl-Meyer Test of Motor Function Adequate cognitive status, as determined by clinical evaluation No upper extremity injury or conditions that limited use prior to the stroke Exclusion Criteria: Cannot give informed consent Have clinically significant fluctuations in mental status within a month of enrollment Were not independent prior to the stroke as measured by scores <95 on the Barthel Index or >1 on the Modified Rankin Scale Have hemispatial neglect as determined by >3 errors on the Star Cancellation Test Have severe sensory loss as determined by a score of 2 on the sensory item of the NIHSS Receiving oral or injected antispasticity medications during study treatment Pain that interferes with daily activities History of prior stroke