PREDICTING TARGETS OF HUMAN REACHING MOTIONS WITH AN ARM REHABILITATION EXOSKELETON

摘要

Rehabilitation robots physically support patients during exercise, but their assistive strategies often constrain patients byforcing them to execute predefined motions. To allow more freedom during rehabilitation, the robot should be able to predictwhat motion the patient wants to perform, then intelligently support the motion. As a first step, this paper presents analgorithm that predicts targets of reaching motions made with an arm rehabilitation exoskeleton. Different sensing modalitiesare compared with regard to their predictive abilities: arm kinematics, eye tracking, contextual information, and combinationsof these modalities. Supervised machine learning is used to make predictions at different points of time during the motion.Results of offline crossvalidation using 12 healthy subjects show that eye tracking can make target predictions earlier andmore accurately than arm kinematics, especially when possible targets are close together. Combining eye tracking withcontextual information further improves prediction accuracy. The foreseen next step is to use our predictions to guide therehabilitation robot, and then test the algorithm in real-time with stroke patients.