The Research Of Haptic Robot System With Active Assistance For Home-based Stroke Rehabilitation

Post-stroke patients may suffer from some loss in motor function and thus need a rehabilitation process to recover from the loss.Stroke patients need to engage themselves in regular,repetitive and intensive rehabilitation exercises for functional recovery.However,such a repetitive exercise is often accompanied by lack of motivation,which may significantly affect the outcome of rehabilitation.It has been fouond that the positive factors affecting the intrinsic motivation are mainly the improvement of the patient's motor function and the support from family and encouragement from partners.Considereing the first factor above the thesis established a new active assistance system with a haptic virtual environment.In this system,a limited active assistance is applied to the patients while they are performing a task to help them complete part of the exercise.The study hypothesized that a proper active assistance with the robotic system increases the motivation of the patients in rehabilitation.The overall objective of the thesis is to verify the above hypothesis and the specific objectives include:(1)developing an active assistance system with a haptic virtual environment for home-based stroke rehabilitation,(2)developing an automatic assessment classifier of the wrist movement function in a haptic virtual environment for home-based stroke rehabilitation,and(3)performing the experiment to determine the effect of the active assistance on intrinsic motivation.The study has concluded:(1)the active assistance provided by the rehabilitation system can affect the intrinsic motivation of patients,and the patient's confidence in completing the task becomes stronger with increase of the assistant level,while the task intensity is reduced by the patient on his own efforts,(2)the accuracy of the motor function assessment model established by BP neural network achieves 98.68%,which is promising in home-based rehabilitation.The main contributions of the thesis are that(1)the wrist rehabilitation was used as an example to answer the question of how the active assistance of the rehabilitation system affects patients' intrinsic motivation,and(2)provided a new motor function assessment model established by BP neural network with highest classification accuracy.