Intelligent robotic rehabilitation system for diagnosis and therapy

In this thesis, it is aimed to develop an intelligent robotic rehabilitation system to assist doctors and physiotherapists in the diagnosis and treatment of wrist and forearm rehabilitation. In this context, a unique robotic system capable of performing biomechanical measurements and conventional therapeutic exercises which were passive, active assistive, stretching, isometric, isotonic, resistive and vario - resistive with variable resistance according to joint range of motion was designed and manufactured. It is possible to treat two patients via this system at the same time. For the control of the system, an artificial intelligence based control structure has been developed that uses information and rule-based algorithms using physical features (age, gender, height, weight, arm peripheral measurements) and biomechanical measurements (grasping force, range of motions, joint force/torque) of patients to identify the lack of joint range of motion and joint force/torque values. According to these deficiencies, the controller can suggest exercise type and parameters to the doctors and can perform therapeutic exercises with conventional control methods. The system has a mobile application that communicates between the patient, the doctor and the robot. With this application, doctors can remotely control the rehabilitation process. As a result of tests with healthy and patient volunteers, it has been shown that the developed intelligent robotic rehabilitation system can perform both diagnosis and treatment in physical therapy and rehabilitation of the wrist and forearm. The contribution of this study to literature is that it can diagnose and treat upper limbs with the developed novel intelligent controller structure and robotic mechanism.