Cognitive-motion Rehabilitation Medical Robot Application Design

People with brain nerve injury and cognitive decline have reduced physical exercise, degraded speech function, and increased risk of falling. The current functional rehabilitation robots mainly focus on passive training, with poor sports rehabilitation performance. Therefore, we studiy a new type of cognitive sports rehabilitation robot and develop its "perception-cognition-motion" multi-modal intelligent perception interaction function. The Pepper robot is used to build a rehabilitation robot system that integrates cognitive and sports rehabilitation training. First, a multi-modal intelligent perception and interaction system is designed to realize the patient's audio-visual tactile interaction and speech training. Next, the robot uses Sonar, laser, and vision to realize real-time map construction and human tracking. Finally, the deep learning method is applied to the robot to achieve human posture recognition based on robot vision, which provides a feasible method for judging the stability of the human body through video images. The experimental results show that the developed rehabilitation robot system positively affects the process of cognitive sports collaborative rehabilitation training. The research proves the feasibility of intelligent robots in medical-assisted rehabilitation training and is expected to be used for home and clinical cognition during synergistic rehabilitation exercises.