Estimation of kinematics from inertial measurement units using a combined deep learning and optimization framework
Introduction
The ability to passively estimate movement kinematics in natural environments with inertial measurement units (IMUs) could transform how we monitor, diagnose, and treat mobility limitations. These devices are now unobtrusive, can flex and bend with the skin, and allow for patient monitoring throughout the day (Patel et al., 2012, Shull et al., 2014, Shull and Damian, 2015, Son et al., 2014). Despite recent progress in hardware miniaturization, turning large multimodal data from wearable sensors into meaningful biomechanical outcomes that can be easily interpreted in the context of healthy and pathological movement remains a key challenge toward their widespread use (Picerno, 2017). Biomechanists and rehabilitation specialists have traditionally characterized gait using joint kinematics and kinetics. While segment inertial data generated by IMUs may also lead to important actionable insights, accurate estimation of joint angles is needed to place future findings in the context of past work.
There are currently no streamlined tools to accurately estimate three-dimensional (3-D) joint kinematics from wearable sensors worn in uncontrolled environments. Strap-down integration of inertial data introduces drift, and sensor fusion algorithms that rely on magnetometer data suffer from ferromagnetic disturbances (de Vries et al., 2009). Solutions that incorporate full-body biomechanical models (Robert-Lachaine et al., 2017a, Robert-Lachaine et al., 2017b, Robert-Lachaine et al., 2020) are currently not portable for anytime, anywhere use, and accuracy over long durations remains to be demonstrated. Additionally, the dependence of most algorithms on accurate sensor-to-segment alignment makes translation difficult for multi-day monitoring outside of the laboratory, given human error in sensor placement. Static and dynamic calibrations (Cutti et al., 2010, Favre et al., 2009, Picerno et al., 2008, Roetenberg et al., 2009) may also add to lack of compliance and increased drop-out rates in remote monitoring studies. Further, static poses or functional calibration trials may not be performed as expected in remote scenarios, especially when users suffer from a mobility-limiting condition.
Deep learning and iterative optimization techniques offer a new opportunity to overcome the limitations of previously proposed approaches for estimating kinematics from IMUs. Deep neural networks are highly efficient in learning non-linear relationships from high-dimensional data, such as dense time series from wearable sensors. A key drawback, however, is that they require large datasets to generate accurate models, and such datasets are scarce in the field of biomechanics. A large dataset that contains both IMU data and ground truth joint kinematics from marker-based motion tracking systems, for example, is currently not available to the research community. In other domains, however, synthetic data have been successfully used to train accurate predictive models (Jaderberg et al., 2014). Additionally, optimization approaches have demonstrated success in improving pose estimation in computer vision applications (von Marcard et al., 2018).
The goal of this study was to build deep learning models for predicting 3-D lower extremity joint kinematics directly from angular velocity and linear acceleration data, circumventing the limitations of magnetometer-dependent algorithms. To generate sufficient data to train such models, we created synthetic inertial data from a marker-based motion capture database of hundreds of subjects collected at a clinical center (Ferber et al., 2014, Osis et al., 2015). We further incorporated data augmentation strategies (Shorten and Khoshgoftaar, 2019) to increase the effective sensor placement variability represented in the data, allowing the developed models to tolerate placement ambiguity without compromising accuracy. Additionally, we used an iterative top-down optimization approach to improve the predictions of the deep learning models.
Section snippets
Data collection and pre-processing
To train the models, we used marker-based motion capture data that were previously collected at the University of Calgary Running Injury Clinic after receiving approval from the University of Calgary’s Conjoint Health Research Ethics Board (Ferber et al., 2016, Jauhiainen et al., 2020, Phinyomark et al., 2018, Pohl et al., 2010). Retro-reflective marker trajectories were collected at 200 Hz using eight high-speed infrared video cameras (Vicon Motion Systems Ltd., Oxford, UK). After a static
Results
Walking kinematics could be predicted with a mean (±STD) RMSE of less than 2.75° (±0.66°), while running kinematics could be predicted with a mean RMSE of less than 3.34 (±1.02°) using the optimized model, along with pseudo-calibration. During walking, flexion/extension was the most accurate degree of freedom, with a mean RMSE of less than 0.97° (±0.38°) across the ankle, knee, and hip joints, followed by ab/adduction with a mean RMSE of less than 2.16° (±0.85°) and internal/external rotation
Discussion
The goal of this study was to introduce a new framework that combines deep neural networks with top-down optimization for prediction of lower extremity kinematics from inertial sensing data. Using this hybrid approach and synthetic inertial data from a large clinical motion capture database, we could predict walking and running kinematics with accuracies that are similar to the reliability of marker-based motion tracking. We also demonstrated that augmentation techniques that increase effective
Acknowledgement
The authors would like to thank Allan Brett for his assistance with the data transfer and for patiently answering questions as the project progressed.
References (36)
- et al.
Magnetic distortion in motion labs, implications for validating inertial magnetic sensors
Gait Posture
(2009) - et al.
Estimation of gait kinematics and kinetics from inertial sensor data using optimal control of musculoskeletal models
J. Biomech.
(2019) - et al.
Functional calibration procedure for 3D knee joint angle description using inertial sensors
J. Biomech.
(2009) - et al.
Gait Biomechanics in the Era of Data Science
J. Biomech.
(2016) - et al.
Machine learning in human movement biomechanics: Best practices, common pitfalls, and new opportunities
J. Biomech.
(2018) - et al.
Kinematic differences between optical motion capture and biplanar videoradiography during a jump-cut maneuver
J. Biomech.
(2013) 25 years of lower limb joint kinematics by using inertial and magnetic sensors: A review of methodological approaches
Gait Posture
(2017)- et al.
Joint kinematics estimate using wearable inertial and magnetic sensing modules
Gait Posture
(2008) - et al.
Can the reliability of three-dimensional running kinematics be improved using functional joint methodology?
Gait Posture
(2010) - et al.
A kinematic and kinetic comparison of overground and treadmill walking in healthy subjects
Gait Posture
(2007)
Accuracy and repeatability of single-pose calibration of inertial measurement units for whole-body motion analysis
Gait Posture
Inertial motion capture validation of 3D knee kinematics at various gait speed on the treadmill with a double-pose calibration
Gait Posture
Quantified self and human movement: A review on the clinical impact of wearable sensing and feedback for gait analysis and intervention
Gait Posture
Effects of soft tissue artifacts on the calculated kinematics and kinetics of the knee during stair-ascent
J. Biomech.
ISB recommendations for standardization in the reporting of kinematic data
J. Biomech.
Estimation of spatio-temporal parameters for post-stroke hemiparetic gait using inertial sensors
Gait Posture
Algorithms for hyper-parameter optimization
Advances in Neural Information Processing Systems.
‘Outwalk’: a protocol for clinical gait analysis based on inertial and magnetic sensors
Med. Biol. Eng. Comput.
Cited by (41)
Integrating an LSTM framework for predicting ankle joint biomechanics during gait using inertial sensors
2024, Computers in Biology and MedicineFusion of video and inertial sensing data via dynamic optimization of a biomechanical model
2023, Journal of BiomechanicsLower body kinematics estimation during walking using an accelerometer
2023, Journal of Biomechanics
- 1
Equal contribution