Muscle innervation patterns for human wrist control: Useful biofeedback signals for robotic rehabilitation?

Upper limb robotic rehabilitation technology has recognized the importance of biofeedback signals for optimizing motor recovery. Impaired motor control is typically associated with abnormal muscle activation patterns and such patterns can be revealed by electromyography (EMG). However, before using user-generated EMG signals as biofeedback for controlling and optimizing force output of a robot, it is imperative that the healthy muscle innervation patterns are fully established to guide possible robot-aided rehabilitation protocols. Given the muscular redundancy found in many human limb systems, and given that force generation of human muscles is not linear, it becomes important to understand how synergistic control influences joint force and joint kinematics. It is the purpose of this study to map the electromyographic activation patterns of the major human muscles involved in controlling the hand/wrist system. By means of a robotic exoskeleton three Degrees of Freedom (DoF) were evaluated: wrist flexion-extension and adduction-abduction as well as forearm pronation-supination separately and for different ranges of the workspace. We recorded EMG activity of 7 arm muscles of 4 healthy subjects performing the task and we therefore defined the typical EMG activation patterns of such muscles during robotic training.