Should Object Function Matter during Modeling of Functional Reach-to-Grasp Tasks in Robot-Assisted Therapy?

Abstract

Recent literature support the idea of using an intense, task-oriented, stroke rehabilitation to promote motor learning and cerebral reorganization. Supporting a task-oriented, robot-assisted therapy approach requires better understanding of the components of real tasks and the limitations and benefits of current trajectory models. We set out to understand natural reach-to-grasp kinematics as it relates to various functional bilateral and unilateral tasks so as to better map this information to a robotic reach-to-grasp therapy systems. To do so, we investigated the influence of arm use and object functionality on four reaching kinematics in reach-to-grasp daily living tasks. We compared our results with the minimum jerk trajectory model used in robot-assisted therapy with the goal of understanding how best to support these real movements in a robotic environment. Eight neurologically intact, right handed subjects participated in the motion analysis study. They completed unilateral and bilateral reaching to objects in the same location with the same orientation, and with handles of the same size and shape. We discuss our results in terms of the minimum jerk model, which is typically used in robot-assisted trajectory planning. Our results showed significant differences in peak velocities, movement time and total displacement across tasks and across arm use conditions.