A Machine Learning and Wearable Sensor Based Approach to Estimate External Knee Flexion and Adduction Moments During Various Locomotion Tasks

Joint moment measurements represent an objective biomechanical parameter of knee joint load in knee osteoarthritis (KOA). Wearable sensors in combination with machine learning techniques may provide solutions to develop assistive devices in KOA patients to improve disease treatment and to minimize risk of non-functional overreaching (e.g., pain). The purpose of this study was to develop an artificial neural network (ANN) that estimates external knee flexion moments (KFM) and external knee adduction moments (KAM) during various locomotion tasks, based on data obtained by two wearable sensors. Thirteen participants were instrumented with two inertial measurement units (IMUs) located on the right thigh and shank. Participants performed six different locomotion tasks consisting of linear motions and motions with a change of direction, while IMU signals as well as full body kinematics and ground reaction forces were synchronously recorded. KFM and KAM were determined using a full body biomechanical model. An ANN was trained to estimate the KFM and KAM time series using the IMU signals as input. Evaluation of the ANN was done using a leave-one-subject-out cross-validation. Concordance of the ANN-estimated KFM and reference data was categorized for five tasks (walking straight, 90° walking turn, moderate running, 90° running turn and 45° cutting maneuver) as strong (r ≥ 0.69, rRMSE ≤ 23.1) and as moderate for fast running (r = 0.65 ± 0.43, rRMSE = 25.5 ± 7.0%). For all locomotion tasks, KAM yielded a lower concordance in comparison to the KFM, ranging from weak (r ≤ 0.21, rRMSE ≥ 33.8%) in cutting and fast running to strong (r = 0.71 ± 0.26, rRMSE = 22.3 ± 8.3%) for walking straight. Smallest mean difference of classical discrete load metrics was seen for KFM impulse, 10.6 ± 47.0%. The results demonstrate the feasibility of using only two IMUs to estimate KFM and KAM to a limited extent. This methodological step facilitates further work that should aim to improve the estimation accuracy to provide valuable biofeedback systems for KOA patients. Greater accuracy of effective implementation could be achieved by a participant- or task-specific ANN modeling.

[1]  Frank J. Wouda,et al.  Estimation of Vertical Ground Reaction Forces and Sagittal Knee Kinematics During Running Using Three Inertial Sensors , 2018, Front. Physiol..

[2]  Bernd J Stetter,et al.  Estimation of Knee Joint Forces in Sport Movements Using Wearable Sensors and Machine Learning , 2019, Sensors.

[3]  Eileen Chih-Ying Yang,et al.  3D analysis system for estimating intersegmental forces and moments exerted on human lower limbs during walking motion , 2015 .

[4]  C. Robinson,et al.  The effect of exercise therapy on knee adduction moment in individuals with knee osteoarthritis: A systematic review. , 2015, Clinical biomechanics.

[5]  J. Barrios,et al.  Gait retraining to reduce the knee adduction moment through real-time visual feedback of dynamic knee alignment. , 2010, Journal of biomechanics.

[6]  J. Favre,et al.  A neural network model to predict knee adduction moment during walking based on ground reaction force and anthropometric measurements. , 2012, Journal of biomechanics.

[7]  Richard M. Smith,et al.  The association of external knee adduction moment with biomechanical variables in osteoarthritis: a systematic review. , 2009, The Knee.

[8]  M. Hurley,et al.  Sensorimotor changes and functional performance in patients with knee osteoarthritis , 1997, Annals of the rheumatic diseases.

[9]  Guang-Zhong Yang,et al.  Wearable Sensing for Solid Biomechanics: A Review , 2015, IEEE Sensors Journal.

[10]  Zoe Y. S. Chan,et al.  Immediate and short-term effects of gait retraining on the knee joint moments and symptoms in patients with early tibiofemoral joint osteoarthritis: a randomized controlled trial. , 2018, Osteoarthritis and cartilage.

[11]  S. Majumdar,et al.  Higher Knee Flexion Moment During the Second Half of the Stance Phase of Gait Is Associated With the Progression of Osteoarthritis of the Patellofemoral Joint on Magnetic Resonance Imaging. , 2015, The Journal of orthopaedic and sports physical therapy.

[12]  H. Koopman,et al.  Prediction of ground reaction forces and moments during various activities of daily living. , 2014, Journal of biomechanics.

[13]  Begonya Garcia-Zapirain,et al.  Gait Analysis Methods: An Overview of Wearable and Non-Wearable Systems, Highlighting Clinical Applications , 2014, Sensors.

[14]  E. Mohammadi,et al.  Barriers and facilitators related to the implementation of a physiological track and trigger system: A systematic review of the qualitative evidence , 2017, International journal for quality in health care : journal of the International Society for Quality in Health Care.

[15]  E. Saitoh,et al.  The impact of ankle–foot orthoses on toe clearance strategy in hemiparetic gait: a cross-sectional study , 2018, Journal of NeuroEngineering and Rehabilitation.

[16]  Jean-Francois Esculier,et al.  2016 Patellofemoral pain consensus statement from the 4th International Patellofemoral Pain Research Retreat, Manchester. Part 1: Terminology, definitions, clinical examination, natural history, patellofemoral osteoarthritis and patient-reported outcome measures , 2016, British Journal of Sports Medicine.

[17]  Clare E. Milner,et al.  A kinematic method to detect foot contact during running for all foot strike patterns. , 2015, Journal of biomechanics.

[18]  P. Veltink,et al.  Ambulatory measurement of the knee adduction moment in patients with osteoarthritis of the knee. , 2013, Journal of biomechanics.

[19]  Scott L Delp,et al.  Quantified self and human movement: a review on the clinical impact of wearable sensing and feedback for gait analysis and intervention. , 2014, Gait & posture.

[20]  Michael J. Burke,et al.  Averaging Correlations: Expected Values and Bias in Combined Pearson rs and Fisher's z Transformations , 1998 .

[21]  Timothy D. Lee,et al.  Motor Control and Learning: A Behavioral Emphasis , 1982 .

[22]  Mark de Zee,et al.  Estimation of the Knee Adduction Moment and Joint Contact Force during Daily Living Activities Using Inertial Motion Capture , 2019, Sensors.

[23]  T. Andriacchi,et al.  Knee adduction moment, serum hyaluronan level, and disease severity in medial tibiofemoral osteoarthritis. , 1998, Arthritis and rheumatism.

[24]  Jolanta Pauk,et al.  Estimation of ground reaction forces and joint moments on the basis on plantar pressure insoles and wearable sensors for joint angle measurement. , 2018, Technology and health care : official journal of the European Society for Engineering and Medicine.

[25]  Jurandir Nadal,et al.  Residual analysis of ground reaction forces simulation during gait using neural networks with different configurations , 2015, 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC).

[26]  Annegret Mündermann,et al.  The role of ambulatory mechanics in the initiation and progression of knee osteoarthritis , 2006, Current opinion in rheumatology.

[27]  Karl E Zelik,et al.  Ground reaction force metrics are not strongly correlated with tibial bone load when running across speeds and slopes: Implications for science, sport and wearable tech , 2019, PloS one.

[28]  Antonie J. van den Bogert,et al.  Authors' reply regarding "Effect of low pass filtering on joint moments from inverse dynamics: Implications for injury prevention" , 2012 .

[29]  Jonathan P. Walter,et al.  Decreased Knee Adduction Moment Does Not Guarantee Decreased Medial Contact Force During Gait , 2009 .

[30]  Andrea Ancillao,et al.  Indirect Measurement of Ground Reaction Forces and Moments by Means of Wearable Inertial Sensors: A Systematic Review , 2018, Sensors.

[31]  Antonie J van den Bogert,et al.  Estimation of gait kinematics and kinetics from inertial sensor data using optimal control of musculoskeletal models. , 2019, Journal of biomechanics.

[32]  Stephen A. Billings,et al.  A New Proxy Measurement Algorithm with Application to the Estimation of Vertical Ground Reaction Forces Using Wearable Sensors , 2017, Sensors.

[33]  Yu Liu,et al.  Lower extremity joint torque predicted by using artificial neural network during vertical jump. , 2009, Journal of biomechanics.

[34]  Madalina Fiterau,et al.  Machine learning in human movement biomechanics: Best practices, common pitfalls, and new opportunities. , 2018, Journal of biomechanics.

[35]  J H van Dieën,et al.  Estimating 3D L5/S1 moments and ground reaction forces during trunk bending using a full-body ambulatory inertial motion capture system. , 2016, Journal of biomechanics.

[36]  Hiroaki Hobara,et al.  Elite long jumpers with below the knee prostheses approach the board slower, but take-off more effectively than non-amputee athletes , 2017, Scientific Reports.

[37]  David Felson,et al.  Association between radiographic features of knee osteoarthritis and pain: results from two cohort studies , 2009, BMJ : British Medical Journal.

[38]  D. Gouwanda,et al.  ANN for gait estimations: A review on current trends and future applications , 2016, 2016 IEEE EMBS Conference on Biomedical Engineering and Sciences (IECBES).

[39]  Mark de Zee,et al.  Estimation of Ground Reaction Forces and Moments During Gait Using Only Inertial Motion Capture , 2016, Sensors.

[40]  Peter H. Veltink,et al.  Influence of the instrumented force shoe on gait pattern in patients with osteoarthritis of the knee , 2011, Medical & Biological Engineering & Computing.

[41]  Richard Taylor.  Interpretation of the Correlation Coefficient: A Basic Review , 1990 .

[42]  F. Cicuttini,et al.  Higher dynamic medial knee load predicts greater cartilage loss over 12 months in medial knee osteoarthritis , 2011, Annals of the rheumatic diseases.

[43]  Josien C van den Noort,et al.  Gait Retraining With Real-Time Biofeedback to Reduce Knee Adduction Moment: Systematic Review of Effects and Methods Used. , 2017, Archives of physical medicine and rehabilitation.

[44]  A J van den Bogert,et al.  Muscle coordination and function during cutting movements. , 1999, Medicine and science in sports and exercise.

[45]  F. Bowling,et al.  Conservative biomechanical strategies for knee osteoarthritis , 2011, Nature Reviews Rheumatology.

[46]  T. Härtel,et al.  Biomechanical modelling and simulation of human body by means of DYNAMICUS , 2006 .

[47]  A. Zaninelli,et al.  Osteoarthritis: an overview of the disease and its treatment strategies. , 2005, Seminars in arthritis and rheumatism.

[48]  Athanasios V. Vasilakos,et al.  Neural networks for computer-aided diagnosis in medicine: A review , 2016, Neurocomputing.

[49]  D. Howard,et al.  Whole body inverse dynamics over a complete gait cycle based only on measured kinematics. , 2008, Journal of biomechanics.

[50]  H Martin Schepers,et al.  Validation of wearable visual feedback for retraining foot progression angle using inertial sensors and an augmented reality headset , 2018, Journal of NeuroEngineering and Rehabilitation.