A Free Placement Approach to Upper-Limb Tracking Using Inertial Sensors

Currently, functional electrical stimulation is used as the most widely used rehabilitation therapy in intelligent medical treatment of stroke patients' limb rehabilitation. There are three methods for tracking optimization of FES rehabilitation therapy using MEMS inertial measurement units: based on Kalman filter algorithm, using quaternion fusion algorithm and establishing sensor information fusion algorithm, but there are existing, calibration needs to be performed in the early stage, inertial measurement unit Must be placed parallel to the arm, noise is prone to data drift and other issues. To solve the problem of the previous calibration of the sensor and the need to place the sensor according to the reference, and to improve the accuracy of the data, we propose a motion tracking model based on free placement to track the upper limb movement trajectory. The input data of the model is the sensor's triaxial acceleration data. The output data is angle change data, which avoids the process of pre-calibration and improves the accuracy of the data through the angle segmentation model. Four sets of comparative experiments show that this method effectively reduces the influence of the sensor position on the data, improves the data accuracy, and has higher accuracy and use value.