Rhythmic Extended Kalman Filter for Gait Rehabilitation Motion Estimation and Segmentation

Abstract

This paper proposes a method to enable the use of non-intrusive, small, wearable, and wireless sensors to estimate the pose of the lower body during gait and other periodic motions and to extract objective performance measures useful for physiotherapy. The Rhythmic Extended Kalman Filter (Rhythmic-EKF) algorithm is developed to estimate the pose, learn an individualized model of periodic movement over time, and use the learned model to improve pose estimation. The proposed approach learns a canonical dynamical system model of the movement during online observation, which is used to accurately model the acceleration during pose estimation. The canonical dynamical system models the motion as a periodic signal. The estimated phase and frequency of the motion also allow the proposed approach to segment the motion into repetitions and extract useful features, such as gait symmetry, step length, and mean joint movement and variance. The algorithm is shown to outperform the extended Kalman filter in simulation, on healthy participant data, and stroke patient data. For the healthy participant marching dataset, the Rhythmic-EKF improves joint acceleration and velocity estimates over regular EKF by 40% and 37%, respectively, estimates joint angles with 2.4° root mean squared error, and segments the motion into repetitions with 96% accuracy.

Inertial Sensor-Based Lower Limb Joint Kinematics: A Methodological Systematic Review

“…Only four out of the thirty-one included studies involved more than ten participants [16][17][18][19]. Furthermore, a predominantly healthy population was recruited, except for identified populations with ankle osteoarthritis [17], transfemoral amputation [20], incomplete spinal cord injury [21], children with cerebral palsy (CP) [19] and stroke patients [22].…”

“…Among the thirty-one studies included in this review, gait was the most commonly evaluated activity, which was assessed either on a treadmill [16,17,23,24] or a walkway [20][21][22][25][26][27][28][29][30][31][32][33][34][35][36][37][38]. Functional movements (i.e., sit-to-stand, squat) are the next most common activities [18,19,32,[39][40][41].…”

“…Furthermore, seven studies focused on more dynamic types of locomotion such as common daily activities [24] and sports, including ski racing [42], running [38,43,44], standing long jump [45] and cycling [46]. Only nine studies reported kinematic analysis beyond 30s of recording [17,21,22,24,29,[42][43][44]46].…”

“…Nearly all the studies made use of sensors placed on adjacent segments surrounding the joint of interest. Four studies opted for a more distal placement on the segments [18,22,35,37]. Only three out of thirty-one studies were able to provide a reproducible description of sensor placement based on anatomical landmarks [19,38,42].…”

“…Only three out of thirty-one studies were able to provide a reproducible description of sensor placement based on anatomical landmarks [19,38,42]. Most of the studies only used three-axes gyroscopes and three-axes accelerometers [17][18][19][20]22,25,26,30,31,33,34,[37][38][39]42,44]. Six studies additionally incorporated a magnetometer [24,29,32,43,45,46].…”