Comparison of measured and predicted reaching movements with a robotic rehabilitation device

Abstract

As has been frequently observed, single-joint movements are characterized by single-peaked, bell-shaped speed profiles. This findings and the tendency of natural movements to be characteristically smooth and graceful led to suggest that motor coordination can be mathematically modeled by postulating that voluntary movements are made, at least in the absence of any other overriding concerns, to be as smooth as possible. Particularly the goal of the central nervous system (CNS) is to maximize smoothness and one measure of this feature is the integrated mean squared magnitude of jerk. Although in the past have been explored kinematic characteristics of reaching movements recorded with actigrafi or goniometric systems and potentiometers, few studies have evaluated the kinematic characteristics of these movements obtained with the new rehabilitative technologies, such as robot mediated therapy. In this paper we tested the applicability of the minimum-jerk model proposed to one join goal directed horizontal reaching movements performed by healthy subjects with a robotic shoulder rehabilitation device. Results show a good qualitative agreement between the measured trajectories and the predicted ones by the model.