Development of control algorithms to enable the implementation of robot-assisted hand rehabilitation therapy modes

A robotic device was developed for characterizing hand function and delivering rehabilitation therapy to the hand and fingers. The current prototype is a five-bar mechanism that interacts with the fingers as they extend/flex in a plane perpendicular to the palm of the hand. This device can simultaneously measure the motion of the fingers and the corresponding forces acting on the fingertips. The mechanism can also assist, resist and control the motion of the fingertips using variable levels of force throughout normal range of motion of the fingers. This doctoral work focuses on the development and validation of control algorithms that will enable several modes of rehabilitation therapy. These algorithms will enable the robotic device to be used as an active or passive hand evaluation tool, an active-resistive exercise therapy tool, and an active-assistive rehabilitation tool. Methods were developed to generate patient-specific paths along which the robot will move the patient's fingers, and several algorithms were created to allow theses paths to be safely repeated at various speeds and for as many cycles as desired. These methods were implemented on the rehabilitation robot and their performance was successfully validated using a group of healthy volunteers.