| Hand injuries are one of the most common body part injuries,accounting for 6.6% to28.6% of all injuries.The main manifestations of hand injuries are joint motion disorders,spasticity,and tremor,which seriously affect the patient’s ability to perform daily activities.The effectiveness of traditional rehabilitation training is easily influenced by the experience and skill level of the therapist,and there is no objective way to measure the severity of motor function loss and changes in functional training.Therefore,this paper proposes a "trainingassessment" integrated soft glove,which provides a safe,reliable,and objective solution to the above-mentioned challenges in rehabilitation therapy.The main research elements are as follows:(1)Based on the theories of human hand structure and kinematic characteristics,the structure and overall scheme of the soft glove were determined.The outer glove provided support for the flexible actuators and assists in hand rehabilitation training;the inner glove was embedded with wearable sensors to provide a data basis for training and assessment.(2)The hand rehabilitation training system was designed,including a pneumatic actuation module,sensor module,and feedback control.The pneumatic actuators were used to improve the flexibility of the soft glove and to realize three modes of rehabilitation training,with 13 training movements.A simple,clear,and easy-to-operate human-computer interaction interface was designed,including four modules: user information,rehabilitation training,functional assessment,and history.It not only visualizes the rehabilitation process and improves patients’ motivation and participation but also effectively transforms patients’ willingness to rehabilitate into the behavior of the soft glove.The safety,reliability,and feasibility of the rehabilitation training system were verified through experiments.The safety assessment results showed that the error between the actual and target motion thresholds of the soft glove was within ±6.5°;the reliability assessment yielded high intra-class correlation coefficient(ICC)(0.7763-0.9996);the experimental results of the multi-modal rehabilitation training showed that the soft glove was able to achieve single-finger separation movements,hand group flexion and extension movements,and inter-finger coordination movements.(3)The hand function assessment system was designed to assess finger motility and hand tremor.The Flex4.5" bending sensors were selected to obtain finger bending signals,and the moving average filtering algorithm was used for noise removal.With the assistance of normalization and K-means clustering analysis,we combined the Total Active Motion(TAM)scale to assess finger motor ability.The experimental results showed that the finger motor ability assessment has high reliability(ICC value of 0.976)and can accurately quantify the finger range of motions,which provides an objective reference for the development of rehabilitation training programs.The MPU9250 sensor was selected to measure the hand tremor signals,and the median,band-pass,and Kalman filtering algorithms were used to perform the denoising process.The tremor signal feature values were extracted by time domain and frequency domain analysis,and the performance of four different machine learning algorithms to achieve tremor classification(No tremor,Mild tremor,and Severe tremor)was compared and analyzed in combination with the sample labels given by the physicians using the Unified Parkinson’s Disease Rating Scale(UPDRS).The experimental results showed that the accuracy of the BP neural network model for classification was as high as 95.83%.Repeated measurement experiments confirmed the high reliability of hand tremor assessment(ICC value of 0.946).In summary,the "training-assessment" integrated soft glove designed in this paper has shown good performance.In practical scenarios,it successfully demonstrates the ability to use objective data to guide rehabilitation training and quantify hand dysfunction,which meets the current clinical rehabilitation needs and solves the problems of personalized hand rehabilitation training,safety,and objectivity of functional assessment. |
