Sensor-based robotic learning system for the rehabilitation of arm
In recent studies, the neurological studies has shown that the neurological disorders like Stroke, Parkinson's disease, Dystonia, Spinal Cord Injury (SCI), Cerebral Palsy (CP), etc., cause the disability of upper limb and one of the major issues with the neurological injuries is motor impairment. This issue is caused by the lack of interaction between the central nervous system (CNS) and muscles  and can be cured by the intensive training which includes repetitive and reaching tasks. The intensive training can be performed using conventional therapy and robot-assisted therapy, but the research has shown additional effects on motor recovery and functional outcome in comparison with conventional treatment forms . The robots which are already developed at different laboratories have achieved in intensively training the neurologically injured patients and have shown remarkable improvement in motor and functional recovery . When assessment is added to the intensive training, the therapists say that the treatment techniques can be optimized . Hence, the assessments were added to the repetitive training by collecting the sensors' (surface EMG (sEMG), accelerometer, and goniometer) data at Laboratory for Industrial Metrology and Automation (LIMA). The results were observed for different modes of operation for 22 healthy subjects, in which 12 were male subjects and 10 were female subjects. This thesis presents the implementation of the robot which has the ability of performing reaching and repetitive tasks; the assessment of the subjects' performance by acquiring the sensors' data in different modes of operation. The modes of operation performed were: No Assistance and No Resistance (NANR) mode, Assistance mode and Resistance mode. In NANR mode, the motor was in the disabled state and the subject performed the experiment in self-selected pace; in assistance mode, the subject was assisted by the motor in two different speeds; and in resistance mode, the motor resisted the movement of the subject in two different torque levels. While the subjects were performing the experiments, the sensors' data was collected and later analyzed to build the knowledge base, where the reference was created. *Please refer to dissertation for references/footnotes.
Keladi, Chethan Ramachandra, "Sensor-based robotic learning system for the rehabilitation of arm" (2012). ETD Collection for University of Texas, El Paso. AAI1512584.