Enhancing Automatic Inertial Sensor Calibration Algorithm for Accurate Joint Angle Estimation in High Flexion Postures

Inertial measurement units (IMUs) offer an appealing solution to the measurement of joint kinematics across a wide range of settings; however, the interpretation of their data requires complex alignment between these sensors and the anatomical axes of each joint. This study sought to evaluate the Seel joint axis algorithm and the proposed Iterative Seel spherical axis (ISSA) extension to this algorithm for sensor to segment alignment and the estimation of ankle and hip angles in the sagittal plane during high flexion movements eliciting increased soft tissue movement in the thigh and shank which are commonly adopted occupationally. These algorithms were validated in nine-such movements as well as in gait across fifty participants. Mean RMSE between the ISSA algorithm extension and optical protocols were 6.61° ± 2.96° and 14.64° ± 6.73° for the ankle and hip respectively. The ISSA algorithm-based estimates however demonstrated strong correlations with gold-standard optical-based angles for both joints in all high flexion movements except for moderate correlations reported for the ankle when sitting on a stool or child-sized chair. The proposed extension to the algorithm is recommended when measuring postures which may elicit high flexion angles and increased soft tissue movement.