Development of VR platform for cloud-based neurorehabilitation and its application to research on sense of agency and ownership

摘要

Recently, neurorehabilitation that uses virtual reality systems is being applied in clinical settings to deal with issues such as phantom limb pain (PLP) as an alternative to mirror box therapy. One of the weak points of mirror box therapy is that the desired analgesia effect might not be confirmed in some patients. One hypothesis to explain this phenomenon is that the subjective sense of the length of a phantom limb is different from that of an intact limb. Since the gap between body representation in the brain and actual sensory feedback is considered one of the causes of PLP, different lengths of a subjective phantom limb are a serious problem for mirror box therapy and similar VR-based rehabilitation methods. We are thus developing a VR system that displays an avatar that has the same length as the subjective phantom limb. The purpose of the current study is to determine the feasibility of the VR system - specifically, whether it has enough effect on sense of agency (SoA) and sense of ownership (SoO) for healthy subjects - before conducting experiments for actual phantom limb patients. To this end, we developed a VR system in which a virtual avatar performs a motion identical to that of the subject by means of a motion capturing device (Kinect V2). The subject wears a 3D head mounted display (Oculus Rift DK2) to experience seeing through the eyes of the avatar. Six conditions of avatar representation were used: two appearances of a normal human arm and a robot arm and three lengths of the arm (short, medium, and long). The subject executes elbow flexion-extension movement of the right arm, which causes the same movement in the VR avatar's arm. After the induction movement, the subjective sense of the length of the right arm is measured by a pointing gesture of the left hand. Twelve subjects participated in this experiment. Results showed that the subjective length of the arm was changed according to the length of the displayed arm in the VR environment. From the results of a questionnaire, we found that there is no negative effect on SoA. SoO when the subjects watch the natural human avatar is stronger than when the robot arm is shown. These results are positive, thus confirming the basic potential of the proposed VR system. In conclusion, the change of self-body appearance of a VR avatar has enough effect on subjective sense of arm length. Since the subjective sense of arm length is strongly related to body representation in the brain, we believe that the system can be a platform for research on embodied-brain science systems.