Multimodal Upper Limb Rehabilitation Training And Evaluation System Based On Cloud-End Integrated

Rehabilitation training and evaluation play an important role in the treatment of limb dysfunction.With the increasing degree of population aging in Chinese society,the number of patients with limb dysfunction is increasing,which makes the research on efficient and low-cost rehabilitation training and evaluation system more urgent.However,traditional rehabilitation training and assessment is dominated by physicians’ manual operations in hospitals,which has the following disadvantages: limited resources such as personnel and space,repetitive and boring training process,the effect of doctors’ experience on evaluation results,costly rehabilitation and the difficulties to record and manage data.In view of the above problems,this project focuses on the multimodal upper limb rehabilitation training and evaluation system based on cloud-end integrated.The main work and innovation are as follows.A rehabilitation evaluation method based on upper limb movement tracking and improved DTW algorithm are proposed.Kinect and IMU were used to obtain the information of patients’ upper limb movement,and we improved the DTW algorithm to calculate the index of similarity between patients’ rehabilitation action and standard action template.Experimental results show that the improved DTW algorithm greatly improves the computational efficiency and reduces the computational burden without affecting the recognition accuracy.A virtual rehabilitation training system with vibrotactile feedback is designed.Rehabilitation training is integrated into virtual game scenes to provide patients with tactile,somatosensory,visual,auditory and other multimodel interactions,in order to increase the authenticity and immersion of rehabilitation training interaction and improve patients’ enthusiasm.Through experimental verification,vibrotactile feedback can help users improve the efficiency and quality of rehabilitation training tasks.A miniaturized somatosensory and tactile feedback unit was designed,which integrates the vibration actuator,IMU,Bluetooth communication module and other modules,so as to reduce the burden of patients’ wearing and the motion constraint during rehabilitation training as much as possible.A rehabilitation medical cloud platform based on My SQL and Javaweb was built,which can receive,record,process and display the rehabilitation training and evaluation data uploaded by various rehabilitation terminals,and send feedback information to the terminals.With this platform,the limitation of time and space between patients and physicians is broken,and the patients’ cost of rehabilitation treatment is reduced,so that patients can receive good,professional and personalized rehabilitation guidance even outside the hospital.