Adaptive IMU-based Calibration of the Center of Joints for Movement Analysis: One Case Study

This work proposes a novel real-time estimator of the center of joints using an inertial measurement unit (IMU), able to adapt to variations in the distance to this center during the joint motion. Our proposal, called ArVE, to the best of our knowledge, is the first real-time estimator of the IMU-joint center distance vector based on IMUs. Previous works are off-line and require a complete measure batch to be solved. We use an extended Kalman filter (EKF) to obtain an IMU-joint vector at each time, instead of the already proposed least-squares approximation, which require a previous collection of measurements and thus the distance vector to be constant. Also, the EKF framework avoids the previous stage of signal noise reduction and directly uses the IMU measurements. We carry out two synthetic experiments in order to evaluate the performance of ArVE in fixed joints when the IMU-joint vector remains constant and when this vector changes over time. We compare the performance of ArVE with a method from the State of the Art. ArVE shows an average error of 1.5 mm and 6.0 mm in the fixed and changing IMU-joint vector cases, respectively. Finally, we test ArVE together with the State of the Art versus an optical system in a real scenario obtaining the IMU-hip vector in one case study. Compared with the optical system, our proposal shows promising results in the test.