Parametrization of an exoskeleton for robotic stroke rehabilitation

Abstract

This paper describes a novel exoskeleton focusing on its parametrization and the redesign of the mechanical metacarpophalangeal joint with an integrated sensor. The joint is based on an arc structure with its rotation axis being aligned with the anatomical center of rotation. A Hall effect based linear encoder is integrated into the base-connected arc reading out a multi-pole magnetic strip in the moving arc. The accuracy was evaluated based on the tendon displacement measured by the motor encoder and converted into the respective angle with two different calculation methods. The parametrization adapts the exoskeleton to the patient’s hand size to avoid misalignment without the use of adaption mechanisms. The exoskeleton is parameterized to a stroke patient. The measurement process is described and evaluated, quantitatively, by comparing two data sets and, qualitatively, by examining the visual overlay of the model onto an image of the hand.