Individualized Generation of Upper Limb Training for Robot-assisted Rehabilitation using Multi-objective Optimization

摘要

To alleviate the labor costs required to customizing the content and difficulty of the rehabilitation game for each patient, we propose a method to automatically generate individualized training contents on desktop end-effector rehabilitation robot. By modeling the search of the training motions as finding optimal hand paths and trajectories, we introduce solving the design problem with a multi-objective optimization (MO) solver. Our system is capable of automatically generating various training plans considering the training intensity and dexterity of each joint in the upper limb. In addition, we develop a serious game to display the generated training, which helps motivate the patient in the rehabilitation.