Health monitoring; inertial measurement unit (IMU); knee kinematics; machine learning; personalized models wearable; Features extraction; IMU; Knee; Knee kinematics; Machine-learning; Personalized model; Personalized-model wearable; Surrogate modeling; Time-series analysis; Instrumentation; Electrical and Electronic Engineering
Abstract :
[en] Knee kinematics is a valuable measure of knee joint function. However, collecting that data outside the clinic is difficult, especially with a limited number of wearable sensors and when you only use an ankle-mounted inertial measurement unit (IMU) to estimate knee kinematics. Due to the cyclic nature of gait, it is possible to use machine learning to extract joint angles from only ankle-mounted sensors. This study aimed to use time-series feature extraction and a random forest regressor to generate a person-specific surrogate model for estimating knee joint flexion angles from a single-mounted IMU above the ankle. Optical motion capture (OMC) and inertial data from ten healthy participants walking on a treadmill were collected to create ten personalized surrogate models for estimating right knee flexion angles during gait. An additional ten models were created for a leave-one-out analysis to test the generalisability of the models. Temporal cross validation of the personalized models and a leave-one-out analysis was performed on the selected feature set. The personalized models achieved an average root-mean-square error (RMSE) of 2.45 \pm 0.65 ( R2 of 0.98) compared to a gold-standard OMC. The generalized models achieved an average RMSE of 6.77 \pm 3.38 ( R2 of 0.83) in the leave-one-out analysis. Time-series feature-based personalized surrogate models could be used to accurately estimate knee kinematics by using a single ankle-mounted sensor. However, more data are required to train a generalized model using the presented method.
Disciplines :
Engineering, computing & technology: Multidisciplinary, general & others
Author, co-author :
Yeung, Ted ; The University of Auckland, Auckland, New Zealand
Cantamessa, Astrid ; Université de Liège - ULiège > Aérospatiale et Mécanique (A&M)
Kempa-Liehr, Andreas W. ; The University of Auckland, Auckland, New Zealand
Besier, Thor ; The University of Auckland, Auckland, New Zealand
Choisne, Julie ; The University of Auckland, Auckland, New Zealand
Language :
English
Title :
Personalized Machine Learning Approach to Estimating Knee Kinematics Using Only Shank-Mounted IMU
Publication date :
June 2023
Journal title :
IEEE Sensors Journal
ISSN :
1530-437X
eISSN :
1558-1748
Publisher :
Institute of Electrical and Electronics Engineers Inc.
K. Lebel, P. Boissy, H. Nguyen, and C. Duval, "Inertial measurement systems for segments and joints kinematics assessment: Towards an understanding of the variations in sensors accuracy," Biomed. Eng. OnLine, vol. 16, no. 1, pp. 1-16, May 2017, doi: 10.1186/s12938-017-0347-6.
L. Vargas-Valencia, A. Elias, E. Rocon, T. Bastos-Filho, and A. Frizera, "An IMU-to-body alignment method applied to human gait analysis," Sensors, vol. 16, no. 12, p. 2090, Dec. 2016, doi: 10.3390/s16122090.
T. McGrath, R. Fineman, and L. Stirling, "An auto-calibrating knee flexion-extension axis estimator using principal component analysis with inertial sensors," Sensors, vol. 18, no. 6, p. 1882, Jun. 2018, doi: 10.3390/s18061882.
M. Nazarahari and H. Rouhani, "Semi-automatic sensor-to-body calibration of inertial sensors on lower limb using gait recording," IEEE Sensors J., vol. 19, no. 24, pp. 12465-12474, Dec. 2019, doi: 10.1109/JSEN.2019.2939981.
M. Al Borno et al., "OpenSense: An open-source toolbox for inertialmeasurement-unit-based measurement of lower extremity kinematics over long durations," J. NeuroEng. Rehabil., vol. 19, no. 1, p. 2021, Dec. 2022, doi: 10.1186/s12984-022-01001-x.
C. A. Bailey, T. K. Uchida, J. Nantel, and R. B. Graham, "Validity and sensitivity of an inertial measurement unit-driven biomechanical model of motor variability for gait," Sensors, vol. 21, no. 22, p. 7690, Nov. 2021, doi: 10.3390/S21227690.
A. Baudet et al., "Cross-talk correction method for knee kinematics in gait analysis using principal component analysis (PCA): A new proposal," PLoS ONE, vol. 9, no. 7, Jul. 2014, Art. no. e102098, doi: 10.1371/journal.pone.0102098.
N. P. Brouwer, T. Yeung, M. F. Bobbert, and T. F. Besier, "3D trunk orientation measured using inertial measurement units during anatomical and dynamic sports motions," Scandin. J. Med. Sci. Sports, vol. 31, no. 2, pp. 358-370, Feb. 2021, doi: 10.1111/sms.13851.
S. O. H. Madgwick, A. J. L. Harrison, and A. Vaidyanathan, "Estimation of IMU and MARG orientation using a gradient descent algorithm," in Proc. IEEE Int. Conf. Rehabil. Robot., Zurich, Switzerland, Jun./Jul. 2011, pp. 1-7, doi: 10.1109/ICORR.2011.5975346.
Z. Zhang, W. C. Wong, and J. Wu, "Wearable sensors for 3D upper limb motion modeling and ubiquitous estimation," J. Control Theory Appl., vol. 9, no. 1, pp. 10-17, Feb. 2011, doi: 10.1007/s11768-011-0234-9.
A. Sakai, Y. Tamura, and Y. Kuroda, "An efficient solution to 6DOF localization using unscented Kalman filter for planetary rovers," in Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst., Oct. 2009, pp. 4154-4159, doi: 10.1109/IROS.2009.5354677.
E.-H. Shin and N. El-Sheimy, "An unscented Kalman filter for in-motion alignment of low-cost IMUs," in Proc. PLANS Position Location Navigat. Symp., Oct. 2004, pp. 273-279, doi: 10.1109/ PLANS.2004.1309005.
C. Dindorf, W. Teufl, B. Taetz, G. Bleser, and M. Fröhlich, "Interpretability of input representations for gait classification in patients after total hip arthroplasty," Sensors, vol. 20, no. 16, p. 4385, Aug. 2020, doi: 10.3390/s20164385.
A. W. Kempa-Liehr, J. Oram, A. Wong, M. Finch, and T. Besier, "Feature engineering workflow for activity recognition from synchronized inertial measurement units," in Proc. Pattern Recognit., ACPR Workshops, Auckland, New Zealand, Nov. 2019, pp. 223-231.
E. Halilaj, A. Rajagopal, M. Fiterau, J. L. Hicks, T. J. Hastie, and S. L. Delp, "Machine learning in human movement biomechanics: Best practices, common pitfalls, and new opportunities," J. Biomech., vol. 81, pp. 1-11, Nov. 2018, doi: 10.1016/j.jbiomech.2018.09.009.
R. Argent, S. Drummond, A. Remus, M. O'Reilly, and B. Caulfield, "Evaluating the use of machine learning in the assessment of joint angle using a single inertial sensor," J. Rehabil. Assistive Technol. Eng., vol. 6, Jan. 2019, Art. no. 205566831986854, doi: 10.1177/2055668319868544.
M. Christ, N. Braun, J. Neuffer, and A. W. Kempa-Liehr, "Time series FeatuRe extraction on basis of scalable hypothesis tests (tsfresh-A Python package)," Neurocomputing, vol. 307, pp. 72-77, Sep. 2018, doi: 10.1016/j.neucom.2018.03.067.
A. A. Slater, T. J. Hullfish, and J. R. Baxter, "The impact of thigh and shank marker quantity on lower extremity kinematics using a constrained model," BMC Musculoskelet. Disord., vol. 19, no. 1, pp. 1-10, Nov. 2018, doi: 10.1186/S12891-018-2329-7/FIGURES/5.
A. Wong and R. Vallabh. (2018). Sensor Specification. IMeasureU. Accessed: Feb. 9, 2022. [Online]. Available: https://imeasureu.com/wpcontent/ uploads/2018/05/Sensor_Specification_v1.5.pdf
T. Yeung, "Marker set (46 makers)," Univ. Auckland, Auckland, Tech. Rep., Apr. 2023. [Online]. Available: https://wiki.auckland. ac.nz/display/ABI/Basic+Mocap+Protocols, doi: 10.17608/k6.auckland. 21588141.v1.
T. F. Besier, D. L. Sturnieks, J. A. Alderson, and D. G. Lloyd, "Repeatability of gait data using a functional hip joint centre and a mean helical knee axis," J. Biomech., vol. 36, pp. 1159-1168, Aug. 2003, doi: 10.1016/S0021-9290(03)00087-3.
S. L. Delp et al., "OpenSim: Open-source software to create and analyze dynamic simulations of movement," IEEE Trans. Biomed. Eng., vol. 54, no. 11, pp. 1940-1950, Nov. 2007, doi: 10.1109/TBME.2007.901024.
J. Zhang et al., "The MAP client: User-friendly musculoskeletal modelling workflows," in Proc. Int. Symp. Biomed. Simulation, 2014, pp. 182-192.
D. Bakke and T. Besier, "Shape model constrained scaling improves repeatability of gait data," J. Biomech., vol. 107, Jun. 2020, Art. no. 109838, doi: 10.1016/j.jbiomech.2020.109838.
S. Schreven, P. J. Beek, and J. B. J. Smeets, "Optimising filtering parameters for a 3D motion analysis system," J. Electromyogr. Kinesiol., vol. 25, no. 5, pp. 808-814, Oct. 2015, doi: 10.1016/j.jelekin. 2015.06.004.
A. Cantamessa, "Evaluating outcome following knee arthroplasty using inertial measurement units," M.S. thesis, Université de Liège, Liège, Belgium, 2020. [Online]. Available: https://matheo. uliege.be/handle/2268.2/10555
H. Zarshenas, B. P. Ruddy, A. W. Kempa-Liehr, and T. F. Besier, "Ankle torque forecasting using time-delayed neural networks," in Proc. 42nd Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. (EMBC), Jul. 2020, pp. 4854-4857.
C. Tang, D. Garreau, and U. von Luxburg, "When do random forests fail?" in Proc. Adv. Neural Inf. Process. Syst., vol. 31, 2018, pp. 2987-2997.
T. Yeung, "Best and worst case," Univ. Auckland, Auckland, New Zealand, Nov. 2022. Accessed: Apr. 25, 2023. [Online]. Available: https://auckland.figshare.com/articles/figure/best_and_worst_ case/21588279/1, doi: 10.17608/k6.auckland.21588279.v1.
T. Yeung, "ML knee joint angle estimation errors-General vs personal," Univ. Auckland, Auckland, New Zealand, Nov. 2022. Accessed: Apr. 25, 2023. [Online]. Available: https://auckland.figshare.com/ articles/figure/ML_Knee_joint_angle_estimation_errors_-_General_vs_ Personal/21597144/1, doi: 10.17608/k6.auckland.21597144.v1.
M. Mundt et al., "Estimation of gait mechanics based on simulated and measured IMU data using an artificial neural network," Frontiers Bioeng. Biotechnol., vol. 8, p. 41, Feb. 2020, doi: 10.3389/fbioe.2020.00041.
S. Chen, S. S. Bangaru, T. Yigit, M. Trkov, C. Wang, and J. Yi, "Realtime walking gait estimation for construction workers using a single wearable inertial measurement unit (IMU)," in Proc. IEEE/ASME Int. Conf. Adv. Intell. Mechatronics (AIM), Jul. 2021, pp. 753-758, doi: 10.1109/AIM46487.2021.9517592.
J.-S. Tan et al., "Predicting knee joint kinematics from wearable sensor data in people with knee osteoarthritis and clinical considerations for future machine learning models," Sensors, vol. 22, no. 2, p. 446, Jan. 2022.
W. Teufl, M. Miezal, B. Taetz, M. Fröhlich, and G. Bleser, "Validity of inertial sensor based 3D joint kinematics of static and dynamic sport and physiotherapy specific movements," PLoS ONE, vol. 14, no. 2, Feb. 2019, Art. no. e0213064, doi: 10.1371/journal.pone.0213064.
R. Jain, V. B. Semwal, and P. Kaushik, "Stride segmentation of inertial sensor data using statistical methods for different walking activities," Robotica, vol. 40, no. 8, pp. 2567-2580, Aug. 2022, doi: 10.1017/S026357472100179X.
B. Oubre et al., "A simple low-cost wearable sensor for long-term ambulatory monitoring of knee joint kinematics," IEEE Trans. Biomed. Eng., vol. 67, no. 12, pp. 3483-3490, Dec. 2020.
A. De Brabandere, P. Robberechts, T. O. D. Beeeck, and J. Davis, "Automating feature construction for multi-view time series data," in Proc. ECMLPKDD Workshop Automating Data Sci., 2019, pp. 1-16.
H. Zarshenas, EMG-Informed Estimation of Human Walking Dynamics for Assistive Robots. Auckland, New Zealand: Univ. Auckland, 2022.
C. Monoli, J. F. Fuentez-Perez, N. Cau, P. Capodaglio, M. Galli, and J. A. Tuhtan, "Land and underwater gait analysis using wearable IMU," IEEE Sensors J., vol. 21, no. 9, pp. 11192-11202, May 2021.
T. Seel, J. Raisch, and T. Schauer, "IMU-based joint angle measurement for gait analysis," Sensors, vol. 14, no. 4, pp. 6891-6909, Apr. 2014, doi: 10.3390/s140406891.
