Purpose: Due to the complexity of the upper motor neuron syndrome, functional improvement in the paretic upper limb after stroke continues to be a challenge in neurorehabilitation. Robot-assisted training has been shown to be useful in relearning gait. In order to achieve similar results in the upper limb, an electromechanical arm robot (ARMOR), capable of moving all joints through complex patterns, has been developed.
Material and method: Eight patients following stroke of different etiologies were included in a clinical AB-BA cross-over study comparing ARMOR training with EMG-triggered neuromuscular electrical stimulation (EMG-NMES). Chedoke-McMaster Stroke Assessment, modified Ashworth Scale, goniometry (Neutral-0-Method), dynamometry and Functional Dexterity Test served as outcome measures.
Results: ARMOR training resulted in more improvement of muscle tone (p = 0.004), range of movement (ROM) (p = 0.005) and dexterity, but less improvement of strength, than EMG-NMES. Chedoke-McMaster Stroke Assessment showed improvement of at least one point in shoulder pain and arm and hand activity during ARMOR training, while these values did not change with EMG-NMS. Better results of ARMOR training were achieved in the earlier phase (A1) than in the later phase (A2).
Conclusion: This study demonstrates the positive effect of automatised training with a new electromechanical arm robot (ARMOR), and documents its clinical applicability in the rehabilitation of the paretic upper extremity in stroke patients.