Ambulatory estimation of 3D lower limb gait posture in anatomical coordinate frame using wearable sensor system

Abstract

Knee joint kinematics analysis using an optimal sensor set and a reliable algorithm would be useful in the gait posture analysis. An original approach for ambulatory estimation of knee joint kinematics in anatomical coordinate system was presented, which was composed of physical-sensor difference based algorithm and virtual-sensor difference based algorithm. To test the approach, a wearable monitoring system composed of accelerometers and magnetometers was developed and evaluated on lower limb. The flexion/extension (f/e), abduction/adduction (a/a) and inversion/extension (i/e) rotation angles of the knee joint in the anatomical joint coordinate system were estimated. In this method, since there was no integration of angular acceleration or angular velocity, the result was not distorted by offset and drift. And the three knee joint angles within the anatomical coordinate system were independent of the orders which must be considered when Euler angles were used. Besides, since there were no physical sensors implanted in the knee joint based on the virtual-sensor difference based algorithm, it was feasible to analyze knee joint kinematics with less numbers and kinds of sensors than ever before. Compared with result from the reference system, the developed wearable sensor system was available to do gait analysis with fewer sensors and high degree of accuracy. Copyright (c) 2010 by JSME.