Human Cognition-Based Power-Assisted Free Motion Object Manipulation for Automating Upper Arm Stroke Rehabilitation

We proposed a 1DOF power assist robotic system to be used by stroke patients for upper arm stroke rehabilitation. An object is tied with the power assist robotic system so a human patient can lift it vertically up and lower it down continuously for a long time for rehabilitation practice. The power assist robotic system is designed to provide power assistance to the patient in lifting and lowering the object. The amount of power assistance can be changed that can change haptically perceived weight or heaviness of the object and vice versa. As a result, the stroke patient can adjust the perceived weight or heaviness of the object at different stages of rehabilitation practices considering the strength and ability of the patient to lift and lower the object with the power assist system at different stages of rehabilitation progress. We derived a computational model to compute power assistance levels of the assist system. We applied a psychophysical method to determine power assistance levels for different perceived weights or heaviness of the manipulated object. We conducted an experiment to evaluate the performance of the power assist robotic system based on an innovative evaluation scheme using healthy human subjects. The results revealed that the subjects were able to realize the variations in the power assistance levels due to variations in the perceived weights or heaviness of the manipulated object. A user study results demonstrated satisfactory user interactions with the power assist robotic system. The proposed approach and mechanism can help automate upper arm rehabilitation of stroke patients, which can improve the quality of life (QoL) of patients.