Evaluation of Error-State Kalman Filter Method for Estimating Human Lower-Limb Kinematics during Various Walking Gaits

Abstract

Inertial measurement units (IMUs) offer an attractive way to study human lower-limb kinematics without traditional laboratory constraints. We present an error-state Kalman filter method to estimate 3D joint angles, joint angle ranges of motion, stride length, and step width using data from an array of seven body-worn IMUs. Importantly, this paper contributes a novel joint axis measurement correction that reduces joint angle drift errors without assumptions of strict hinge-like joint behaviors of the hip and knee. We evaluate the method compared to two optical motion capture methods on twenty human subjects performing six different types of walking gait consisting of forward walking (at three speeds), backward walking, and lateral walking (left and right). For all gaits, RMS differences in joint angle estimates generally remain below 5 degrees for all three ankle joint angles and for flexion/extension and abduction/adduction of the hips and knees when compared to estimates from reflective markers on the IMUs. Additionally, mean RMS differences in estimated stride length and step width remain below 0.13 m for all gait types, except stride length during slow walking. This study confirms the method’s potential for non-laboratory based gait analysis, motivating further evaluation with IMU-only measurements and pathological gaits.

Drift-Free Joint Angle Calculation Using Inertial Measurement Units without Magnetometers: An Exploration of Sensor Fusion Methods for the Elbow and Wrist

“…They reported 3.8° RMSE for knee flexion/extension and 3.7°, 6.3°, and 4.6° for hip flexion/extension, internal/external rotation, and abduction/adduction, respectively, for 30 min of walking across 12 participants. Potter et al [ 29 ] reported that RMSEs were generally <5° for all ankle joint angles and for flexion/extension and abduction/adduction of the hips and knee in their study of 20 participants that performed one-minute walking trials, with each trial consisting of different walking speeds as well as backward and lateral walking. Similar to our study, it was observed that RMSEs are joint and motion dependent.…”

“…IMCs can provide drift-free joint angle measurements in all three rotational movement directions without using magnetometers by incorporating linear and/or rotational motion constraints [ 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 ]. Linear motion constraints assume the linear acceleration [ 21 , 30 ], linear velocity [ 31 ], or linear position [ 22 , 25 , 26 , 29 , 32 , 33 ] at the joint center must be equal when determined from either of the two IMUs attached to adjacent body segments.…”

“…IMCs can provide drift-free joint angle measurements in all three rotational movement directions without using magnetometers by incorporating linear and/or rotational motion constraints [ 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 ]. Linear motion constraints assume the linear acceleration [ 21 , 30 ], linear velocity [ 31 ], or linear position [ 22 , 25 , 26 , 29 , 32 , 33 ] at the joint center must be equal when determined from either of the two IMUs attached to adjacent body segments. A rotational constraint assumes that joints contain limited rotational motion in at least one of the three motion planes (e.g., limited abduction/adduction movement of the elbow) [ 27 , 28 , 34 ] or that joints are limited by their range of motion [ 24 , 35 , 36 , 37 ].…”

Gait Analysis in Neurorehabilitation: From Research to Clinical Practice

“…It seems that both solutions may provide detailed and clinically relevant information, although the mobile application was a more ecological and user-friendly solution, calculating stride, swing and stance phase duration, and cadence. Interestingly, Liu et al [32] performed a quantitative analysis using an error state Kalman filter (which is a useful method to estimate the kinematic of lower limb during gait reducing joint angle drift errors) by extrapolating data from WS systems [21]. The authors found that this innovative method can recognize altered gait patterns of PD patients when compared to healthy subjects with similar demographic features, suggesting its use as an objective gait assessment tool.…”

The Use of Head-Mounted Display Systems for Upper Limb Kinematic Analysis in Post-Stroke Patients: A Perspective Review on Benefits, Challenges and Other Solutions

“…These systems are particularly popular since they have several advantages, being low-cost, wearable, easy-to-use, settings which allow continuous monitoring of movement both in the clinic and at home [ 8 ]. IMUs collect raw data from each integrated sensor through on-board sensor-fusion algorithms, based on Kalman filters [ 13 ], in order to obtain different motion parameters. Generally, these systems allow the recording of kinematics, providing data on angular acceleration, velocity, and spatial orientation, from different body parts, including upper limbs [ 8 , 10 ].…”