A rehabilitation gait training system for half lower limb disorder

Gait training is one of main means of rehabilitation of lower limb disfunction. Nevertheless, the promotion of clinical gait training is inhibited by the professional skill and labor consumption of physiatrist. It is with great practical value to design an automatic rehabilitation equipment which could increase the effectiveness and quality of training progress, meanwhile reduce labor costs. In this paper, we presented a new automatic rehabilitation gait training system constructed by wearable sensors and robotic manipulator for hemiplegia patients. After that, using the measurement of normal lower limb, the correction of the disorder lower limb has been estimated. For the detection of the normal lower limb motion, a movement capturing system based on a Kinect and several wearable initial measurement units (IMU) is proposed, and a fusion method is designed to handle time synchronization between multiple sensors and to alleviate cumulative errors. After analyzing the lower limb trajectories, a gait model is constructed based on the regression model with spline interpolation. Finally, corrected lower limb gait trajectories generated by the constructed gait model is simulated using 3D animation, which illustrate the stability and practicability of the proposed system.