{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,28]],"date-time":"2026-06-28T05:41:06Z","timestamp":1782625266895,"version":"3.54.5"},"reference-count":29,"publisher":"MDPI AG","issue":"13","license":[{"start":{"date-parts":[[2021,7,1]],"date-time":"2021-07-01T00:00:00Z","timestamp":1625097600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>The application of artificial intelligence techniques to wearable sensor data may facilitate accurate analysis outside of controlled laboratory settings\u2014the holy grail for gait clinicians and sports scientists looking to bridge the lab to field divide. Using these techniques, parameters that are difficult to directly measure in-the-wild, may be predicted using surrogate lower resolution inputs. One example is the prediction of joint kinematics and kinetics based on inputs from inertial measurement unit (IMU) sensors. Despite increased research, there is a paucity of information examining the most suitable artificial neural network (ANN) for predicting gait kinematics and kinetics from IMUs. This paper compares the performance of three commonly employed ANNs used to predict gait kinematics and kinetics: multilayer perceptron (MLP); long short-term memory (LSTM); and convolutional neural networks (CNN). Overall high correlations between ground truth and predicted kinematic and kinetic data were found across all investigated ANNs. However, the optimal ANN should be based on the prediction task and the intended use-case application. For the prediction of joint angles, CNNs appear favourable, however these ANNs do not show an advantage over an MLP network for the prediction of joint moments. If real-time joint angle and joint moment prediction is desirable an LSTM network should be utilised.<\/jats:p>","DOI":"10.3390\/s21134535","type":"journal-article","created":{"date-parts":[[2021,7,1]],"date-time":"2021-07-01T12:03:27Z","timestamp":1625141007000},"page":"4535","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":95,"title":["A Comparison of Three Neural Network Approaches for Estimating Joint Angles and Moments from Inertial Measurement Units"],"prefix":"10.3390","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6624-2895","authenticated-orcid":false,"given":"Marion","family":"Mundt","sequence":"first","affiliation":[{"name":"Minderoo Tech and Policy Lab, UWA Law School, The University of Western Australia, Crawley 6009, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6401-4597","authenticated-orcid":false,"given":"William R.","family":"Johnson","sequence":"additional","affiliation":[{"name":"Houston Astros Baseball Club, Houston, TX 77001, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9135-3802","authenticated-orcid":false,"given":"Wolfgang","family":"Potthast","sequence":"additional","affiliation":[{"name":"Institute of Biomechanics and Orthopeadics, German Sport University Cologne, 50933 Cologne, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7893-6229","authenticated-orcid":false,"given":"Bernd","family":"Markert","sequence":"additional","affiliation":[{"name":"Institute of General Mechanics, RWTH Aachen University, 52062 Aachen, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ajmal","family":"Mian","sequence":"additional","affiliation":[{"name":"School of Computer Science and Software Engineering, The University of Western Australia, Crawley 6009, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8866-0913","authenticated-orcid":false,"given":"Jacqueline","family":"Alderson","sequence":"additional","affiliation":[{"name":"Minderoo Tech and Policy Lab, UWA Law School, The University of Western Australia, Crawley 6009, Australia"},{"name":"Sports Performance Research Institute New Zealand (SPRINZ), Auckland University of Technology, Auckland 1010, New Zealand"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2021,7,1]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","unstructured":"Adesida, Y., Papi, E., and McGregor, A.H. (2019). Exploring the role of wearable technology in sport kinematics and kinetics: A systematic review. Sensors, 19.","DOI":"10.3390\/s19071597"},{"key":"ref_2","doi-asserted-by":"crossref","unstructured":"Gurchiek, R.D., Cheney, N., and Mcginnis, R.S. (2019). Estimating Biomechanical Time-Series with Wearable Sensors: A Systematic Review of Machine Learning Techniques. Sensors, 19.","DOI":"10.20944\/preprints201911.0006.v1"},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s12984-020-00685-3","article-title":"Validity and reliability of wearable inertial sensors in healthy adult walking: A systematic review and meta-analysis","volume":"17","author":"Kobsar","year":"2020","journal-title":"J. Neuroeng. Rehabil."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"543","DOI":"10.1016\/S0021-9290(01)00222-6","article-title":"ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion\u2014Part I: Ankle, hip, and spine","volume":"35","author":"Wu","year":"2002","journal-title":"J. Biomech."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"109832","DOI":"10.1016\/j.jbiomech.2020.109832","article-title":"Determining anatomical frames via inertial motion capture: A survey of methods","volume":"106","author":"Vitali","year":"2020","journal-title":"J. Biomech."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Caruso, M., Sabatini, A.M., Laidig, D., Seel, T., Knaflitz, M., Croce, U.D., and Cereatti, A. (2021). Analysis of the accuracy of ten algorithms for orientation estimation using inertial and magnetic sensing under optimal conditions: One size does not fit all. Sensors, 21.","DOI":"10.3390\/s21072543"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"145","DOI":"10.1016\/j.measurement.2014.03.004","article-title":"Experimental evaluation of accuracy and repeatability of a novel body-to-sensor calibration procedure for inertial sensor-based gait analysis","volume":"52","author":"Palermo","year":"2014","journal-title":"Meas. J. Int. Meas. Confed."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"80","DOI":"10.1016\/j.gaitpost.2017.02.029","article-title":"Accuracy and repeatability of single-pose calibration of inertial measurement units for whole-body motion analysis","volume":"54","author":"Mecheri","year":"2017","journal-title":"Gait Posture"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Ancillao, A., Tedesco, S., Barton, J., and O\u2019flynn, B. (2018). Indirect measurement of ground reaction forces and moments by means of wearable inertial sensors: A systematic review. Sensors, 18.","DOI":"10.3390\/s18082564"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"1","DOI":"10.3389\/fphys.2018.00218","article-title":"Estimation of Vertical Ground Reaction Forces and Sagittal Knee Kinematics During Running Using Three Inertial Sensors","volume":"9","author":"Wouda","year":"2018","journal-title":"Front. Physiol."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"1","DOI":"10.3389\/fbioe.2020.00604","article-title":"CNN-Based Estimation of Sagittal Plane Walking and Running Biomechanics From Measured and Simulated Inertial Sensor Data","volume":"8","author":"Dorschky","year":"2020","journal-title":"Front. Bioeng. Biotechnol."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Lim, H., Kim, B., and Park, S. (2020). Prediction of Lower Limb Kinetics and Kinematics during Walking by a Single IMU on the Lower Back Using Machine Learning. Sensors, 20.","DOI":"10.3390\/s20010130"},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Stetter, B.J., Krafft, F.C., Ringhof, S., Stein, T., and Sell, S. (2020). A Machine Learning and Wearable Sensor Based Approach to Estimate External Knee Flexion and Adduction Moments During Various Locomotion Tasks. Front. Bioeng. Biotechnol., 8.","DOI":"10.3389\/fbioe.2020.00009"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Mundt, M., Koeppe, A., David, S., Witter, T., Bamer, F., Potthast, W., and Markert, B. (2020). Estimation of Gait Mechanics Based on Simulated and Measured IMU Data Using an Artificial Neural Network. Front. Bioeng. Biotechnol.","DOI":"10.3389\/fbioe.2020.00041"},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Rapp, E., Shin, S., Thomsen, W., Ferber, R., and Halilaj, E. (2021). Estimation of kinematics from inertial measurement units using a combined deep learning and optimization framework. J. Biomech., 116.","DOI":"10.1016\/j.jbiomech.2021.110229"},{"key":"ref_16","unstructured":"Goodfellow, I., Bengio, Y., and Courville, A. (2016). Deep Learning, MIT Press. Available online: http:\/\/www.deeplearningbook.org."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"211","DOI":"10.1007\/s11517-019-02061-3","article-title":"Prediction of lower limb joint angles and moments during gait using artificial neural networks","volume":"58","author":"Mundt","year":"2020","journal-title":"Med. Biol. Eng. Comput."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"689","DOI":"10.1109\/TBME.2018.2854632","article-title":"Predicting Athlete Ground Reaction Forces and Moments From Spatio-Temporal Driven CNN Models","volume":"66","author":"Johnson","year":"2019","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Johnson, W.R., Mian, A., Robinson, M.A., Verheul, J., Lloyd, D.G., and Alderson, J. (2020). Multidimensional ground reaction forces and moments from wearable sensor accelerations via deep learning. IEEE Trans. Biomed. Eng., 1\u201312.","DOI":"10.1109\/TBME.2020.3006158"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Mundt, M., Koeppe, A., Bamer, F., David, S., and Markert, B. (2020). Artificial Neural Networks in Motion Analysis\u2014Applications of Unsupervised and Heuristic Feature Selection Techniques. Sensors, 20.","DOI":"10.3390\/s20164581"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1371\/journal.pone.0168566","article-title":"Non-sagittal knee joint kinematics and kinetics during gait on level and sloped grounds with unicompartmental and total knee arthroplasty patients","volume":"11","author":"Komnik","year":"2016","journal-title":"PLoS ONE"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"73","DOI":"10.1016\/j.jbiomech.2018.12.013","article-title":"Agreement of measured and calculated muscle activity during highly dynamic movements modelled with a spherical knee joint","volume":"84","author":"Dietzsch","year":"2019","journal-title":"J. Biomech."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Mundt, M., Thomsen, W., Bamer, F., and Markert, B. (2018). Determination of gait parameters in real-world environment using low-cost inertial sensors. PAMM, 18.","DOI":"10.1002\/pamm.201800014"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"81","DOI":"10.1016\/j.jbiomech.2018.12.023","article-title":"Assessment of the measurement accuracy of inertial sensors during different tasks of daily living","volume":"84","author":"Mundt","year":"2019","journal-title":"J. Biomech."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Robertson, G., Caldwell, G., Hamill, J., Kamen, G., and Whittlesey, S. (2013). Research Methods in Biomechanics, Human Kinetics. [2nd ed.].","DOI":"10.5040\/9781492595809"},{"key":"ref_26","first-page":"1","article-title":"Hyperband: A novel bandit-based approach to hyperparameter optimization","volume":"18","author":"Li","year":"2018","journal-title":"J. Mach. Learn. Res."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"3279","DOI":"10.1007\/s00707-019-02436-5","article-title":"An efficient Monte Carlo strategy for elasto-plastic structures based on recurrent neural networks","volume":"230","author":"Koeppe","year":"2019","journal-title":"Acta Mech."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"29","DOI":"10.1016\/j.medengphy.2020.10.001","article-title":"Prediction of ground reaction force and joint moments based on optical motion capture data during gait","volume":"86","author":"Mundt","year":"2020","journal-title":"Med. Eng. Phys."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"185","DOI":"10.1016\/j.jbiomech.2019.07.002","article-title":"On-field player workload exposure and knee injury risk monitoring via deep learning","volume":"93","author":"Johnson","year":"2019","journal-title":"J. Biomech."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/21\/13\/4535\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T06:24:50Z","timestamp":1760163890000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/21\/13\/4535"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,7,1]]},"references-count":29,"journal-issue":{"issue":"13","published-online":{"date-parts":[[2021,7]]}},"alternative-id":["s21134535"],"URL":"https:\/\/doi.org\/10.3390\/s21134535","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,7,1]]}}}