A new approach to Microsoft Kinect for neuro-motor rehabilitation of the upper body

Stroke is one of the leading causes of death worldwide and the main factor of neuromotor impairments. Neurorehabilitation plays an important role in limiting and improving this pathological situation. Critical elements of rehabilitation are to provide an engaging therapy, the ability to operate in a domestic environment, at low cost and the capability to assess the patients’ state. The Human Robotics Group at Imperial College London has developed different rehabilitation devices, in particular the SITAR, which is an upper limb rehabilitation system. The SITAR addresses the factors of providing and engaging environment, is suitable for domestic use and inexpensive. It permits the interaction with patients through a large force touch screen table and require them to move objects and make specific movements, and measure the performance. This apparatus, however is incapable of controlling and assessing all the particular movement configurations during the tasks, such as compensatory movements, that impaired individuals tends to do. This thesis shows that coupling the SITAR with Microsoft Kinect can effectively be used to control the rehabilitation process. Through the use of Microsoft Kinect and the developed algorithms to body recognition and motion tracking, it becomes possible to carry out the operations normally done using motion capture device, i.e. retrieving kinematic parameters necessary for assessing movement in neurorehabilitation. The chosen system has been verified through comparison with the gold standard device VICON system.