{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,4]],"date-time":"2026-07-04T02:47:28Z","timestamp":1783133248332,"version":"3.54.6"},"reference-count":42,"publisher":"MDPI AG","issue":"18","license":[{"start":{"date-parts":[[2023,9,6]],"date-time":"2023-09-06T00:00:00Z","timestamp":1693958400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000738","name":"United States Department of Veterans Affairs Rehabilitation Research and Development Service","doi-asserted-by":"publisher","award":["I01 RX003138"],"award-info":[{"award-number":["I01 RX003138"]}],"id":[{"id":"10.13039\/100000738","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000738","name":"United States Department of Veterans Affairs Rehabilitation Research and Development Service","doi-asserted-by":"publisher","award":["I50 RX002357"],"award-info":[{"award-number":["I50 RX002357"]}],"id":[{"id":"10.13039\/100000738","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000738","name":"United States Department of Veterans Affairs Rehabilitation Research and Development Service","doi-asserted-by":"publisher","award":["IK6 RX002974"],"award-info":[{"award-number":["IK6 RX002974"]}],"id":[{"id":"10.13039\/100000738","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>Inverse dynamics from motion capture is the most common technique for acquiring biomechanical kinetic data. However, this method is time-intensive, limited to a gait laboratory setting, and requires a large array of reflective markers to be attached to the body. A practical alternative must be developed to provide biomechanical information to high-bandwidth prosthesis control systems to enable predictive controllers. In this study, we applied deep learning to build dynamical system models capable of accurately estimating and predicting prosthetic ankle torque from inverse dynamics using only six input signals. We performed a hyperparameter optimization protocol that automatically selected the model architectures and learning parameters that resulted in the most accurate predictions. We show that the trained deep neural networks predict ankle torques one sample into the future with an average RMSE of 0.04 \u00b1 0.02 Nm\/kg, corresponding to 2.9 \u00b1 1.6% of the ankle torque\u2019s dynamic range. Comparatively, a manually derived analytical regression model predicted ankle torques with a RMSE of 0.35 \u00b1 0.53 Nm\/kg, corresponding to 26.6 \u00b1 40.9% of the ankle torque\u2019s dynamic range. In addition, the deep neural networks predicted ankle torque values half a gait cycle into the future with an average decrease in performance of 1.7% of the ankle torque\u2019s dynamic range when compared to the one-sample-ahead prediction. This application of deep learning provides an avenue towards the development of predictive control systems for powered limbs aimed at optimizing prosthetic ankle torque.<\/jats:p>","DOI":"10.3390\/s23187712","type":"journal-article","created":{"date-parts":[[2023,9,6]],"date-time":"2023-09-06T10:23:42Z","timestamp":1693995822000},"page":"7712","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":9,"title":["Using Deep Learning Models to Predict Prosthetic Ankle Torque"],"prefix":"10.3390","volume":"23","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5433-7915","authenticated-orcid":false,"given":"Christopher","family":"Prasanna","sequence":"first","affiliation":[{"name":"Center for Limb Loss and Mobility, Seattle, WA 98108, USA"},{"name":"Department of Mechanical Engineering, University of Washington, Seattle, WA 98195, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3318-9770","authenticated-orcid":false,"given":"Jonathan","family":"Realmuto","sequence":"additional","affiliation":[{"name":"Bionic Systems Lab, University of California, Riverside, CA 92521, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Anthony","family":"Anderson","sequence":"additional","affiliation":[{"name":"Center for Limb Loss and Mobility, Seattle, WA 98108, USA"},{"name":"Department of Mechanical Engineering, University of Washington, Seattle, WA 98195, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8523-1913","authenticated-orcid":false,"given":"Eric","family":"Rombokas","sequence":"additional","affiliation":[{"name":"Department of Mechanical Engineering, University of Washington, Seattle, WA 98195, USA"},{"name":"Department of Electrical and Computer Engineering, University of Washington, Seattle, WA 98195, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9527-9590","authenticated-orcid":false,"given":"Glenn","family":"Klute","sequence":"additional","affiliation":[{"name":"Center for Limb Loss and Mobility, Seattle, WA 98108, USA"},{"name":"Department of Mechanical Engineering, University of Washington, Seattle, WA 98195, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2023,9,6]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"319","DOI":"10.1007\/s10339-011-0404-1","article-title":"Model learning for robot control: A survey","volume":"12","author":"Peters","year":"2011","journal-title":"Cogn. Process."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"210","DOI":"10.1109\/LRA.2017.2737478","article-title":"EMG-Based Model Predictive Control for Physical Human\u2013Robot Interaction: Application for Assist-As-Needed Control","volume":"3","author":"Teramae","year":"2018","journal-title":"IEEE Robot. Autom. Lett."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"\u00d6zen, O., Traversa, F., Gadi, S., Buetler, K.A., Nef, T., and Marchal-Crespo, L. (2019, January 24\u201328). Multi-purpose Robotic Training Strategies for Neurorehabilitation with Model Predictive Controllers. Proceedings of the 2019 IEEE 16th International Conference on Rehabilitation Robotics (ICORR), Toroto, ON, Canada.","DOI":"10.1109\/ICORR.2019.8779396"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"51","DOI":"10.1016\/0167-9457(84)90005-8","article-title":"Kinematic and kinetic patterns in human gait: Variability and compensating effects","volume":"3","author":"Winter","year":"1984","journal-title":"Hum. Mov. Sci."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/1743-0003-12-1","article-title":"Control strategies for active lower extremity prosthetics and orthotics: A review","volume":"12","author":"Tucker","year":"2015","journal-title":"J. Neuroeng. Rehabil."},{"key":"ref_6","first-page":"369","article-title":"Generalization in Reinforcement Learning: Safely Approximating the Value Function","volume":"Volume 7","author":"Boyan","year":"1994","journal-title":"Proceedings of the 7th International Conference on Neural Information Processing Systems"},{"key":"ref_7","unstructured":"Atkeson, C. (December, January 30). Using Local Trajectory Optimizers to Speed Up Global Optimization in Dynamic Programming. Proceedings of the Advances in Neural Information Processing Systems, Denver, CO, USA."},{"key":"ref_8","unstructured":"Rawlings, J.B., Mayne, D.Q., and Diehl, M.M. (2017). Model Predictive Control: Theory, Computation, and Design, Nob Hill Publishing. [2nd ed.]."},{"key":"ref_9","first-page":"1057","article-title":"Policy gradient methods for reinforcement learning with function approximation","volume":"12","author":"Sutton","year":"2000","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"345","DOI":"10.1613\/jair.4992","article-title":"A primer on neural network models for natural language processing","volume":"57","author":"Goldberg","year":"2016","journal-title":"J. Artif. Intell. Res."},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Redmon, J., Divvala, S., Girshick, R., and Farhadi, A. (2016, January 27\u201330). You only look once: Unified, real-time object detection. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.","DOI":"10.1109\/CVPR.2016.91"},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"970","DOI":"10.1109\/LRA.2020.2966390","article-title":"Real time trajectory prediction using deep conditional generative models","volume":"5","author":"Prokudin","year":"2020","journal-title":"IEEE Robot. Autom. Lett."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Almeida, A., and Azkune, G. (2018). Predicting human behaviour with recurrent neural networks. Appl. Sci., 8.","DOI":"10.3390\/app8020305"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"542","DOI":"10.1109\/TBME.2009.2034734","article-title":"Multiclass real-time intent recognition of a powered lower limb prosthesis","volume":"57","author":"Varol","year":"2010","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Young, A.J., Simon, A.M., Fey, N.P., and Hargrove, L.J. (2013, January 6\u20138). Classifying the intent of novel users during human locomotion using powered lower limb prostheses. Proceedings of the International IEEE\/EMBS Conference on Neural Engineering, NER, San Diego, CA, USA.","DOI":"10.1109\/NER.2013.6695934"},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"2867","DOI":"10.1109\/TBME.2011.2161671","article-title":"Continuous locomotion-mode identification for prosthetic legs based on neuromuscular-mechanical fusion","volume":"58","author":"Huang","year":"2011","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Vu, H.T.T., Dong, D., Cao, H.L., Verstraten, T., Lefeber, D., Vanderborght, B., and Geeroms, J. (2020). A review of gait phase detection algorithms for lower limb prostheses. Sensors, 20.","DOI":"10.3390\/s20143972"},{"key":"ref_18","unstructured":"Au, S.K., Bonato, P., and Herr, H. (2005, January 28). An EMG-position controlled system for an active ankle-foot prosthesis: An initial experimental study. Proceedings of the 9th International Conference on Rehabilitation Robotics, Chicago, IL, USA."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"654","DOI":"10.1016\/j.neunet.2008.03.006","article-title":"Powered ankle-foot prosthesis to assist level-ground and stair-descent gaits","volume":"21","author":"Au","year":"2008","journal-title":"Neural Netw."},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Dedi\u0107, R., and Dindo, H. (2011, January 27\u201329). SmartLeg: An intelligent active robotic prosthesis for lower-limb amputees. Proceedings of the 2011 23rd International Symposium on Information, Communication and Automation Technologies, ICAT, Sarajevo, Bosnia and Herzegovina.","DOI":"10.1109\/ICAT.2011.6102090"},{"key":"ref_21","unstructured":"Rai, V., Sharma, A., Preechayasomboon, P., and Rombokas, E. (December, January 29). Coordinated movement for prosthesis reference trajectory generation: Temporal factors and attention. Proceedings of the IEEE RAS and EMBS International Conference on Biomedical Robotics and Biomechatronics, New York, NY, USA."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"699","DOI":"10.1109\/TNSRE.2022.3156884","article-title":"Improving IMU-Based Prediction of Lower Limb Kinematics in Natural Environments Using Egocentric Optical Flow","volume":"30","author":"Sharma","year":"2022","journal-title":"IEEE Trans. Neural Syst. Rehabil. Eng."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"111020","DOI":"10.1016\/j.jbiomech.2022.111020","article-title":"Predicting biological joint moment during multiple ambulation tasks","volume":"134","author":"Camargo","year":"2022","journal-title":"J. Biomech."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"011003","DOI":"10.1088\/1741-2552\/ac4f9a","article-title":"Deep learning for biosignal control: Insights from basic to real-time methods with recommendations","volume":"19","author":"Dillen","year":"2022","journal-title":"J. Neural Eng."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"E24","DOI":"10.1017\/wtc.2022.19","article-title":"Seamless and intuitive control of a powered prosthetic leg using deep neural network for transfemoral amputees","volume":"3","author":"Kim","year":"2022","journal-title":"Wearable Technol."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"2829","DOI":"10.1109\/JBHI.2023.3262164","article-title":"Estimation of Lower Extremity Joint Moments and 3D Ground Reaction Forces Using IMU Sensors in Multiple Walking Conditions: A Deep Learning Approach","volume":"27","author":"Hossain","year":"2023","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"219","DOI":"10.1109\/TMRB.2022.3144025","article-title":"Subject-Independent, Biological Hip Moment Estimation During Multimodal Overground Ambulation Using Deep Learning","volume":"4","author":"Molinaro","year":"2022","journal-title":"IEEE Trans. Med. Robot. Bionics"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"85","DOI":"10.1016\/j.neunet.2014.09.003","article-title":"Deep learning in neural networks: An overview","volume":"61","author":"Schmidhuber","year":"2015","journal-title":"Neural Netw."},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Cho, K., Van Merri\u00ebnboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., Bengio, Y., van Merrienboer, B., Gulcehre, C., and Bahdanau, D. (2014, January 25\u201329). Learning phrase representations using RNN encoder-decoder for statistical machine translation. Proceedings of the EMNLP 2014\u20142014 Conference on Empirical Methods in Natural Language Processing, Association for Computational Linguistics (ACL), Doha, Qatar.","DOI":"10.3115\/v1\/D14-1179"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Qin, Y., Song, D., Chen, H., Cheng, W., Jiang, G., and Cottrell, G. (2017, January 19\u201325). A dual-stage attention-based recurrent neural network for time series prediction. Proceedings of the IJCAI International Joint Conference on Artificial Intelligence, Melbourne, Australia.","DOI":"10.24963\/ijcai.2017\/366"},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"011007","DOI":"10.1115\/1.4028653","article-title":"Nonlinear passive cam-based springs for powered ankle prostheses","volume":"9","author":"Realmuto","year":"2015","journal-title":"J. Med. Devices Trans. ASME"},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Realmuto, J., Klute, G., and Devasia, S. (2019, January 25\u201327). Preliminary investigation of symmetry learning control for powered ankle-foot prostheses. Proceedings of the 2019 Wearable Robotics Association Conference (WearRAcon), Scottsdale, AZ, USA.","DOI":"10.1109\/WEARRACON.2019.8719630"},{"key":"ref_33","unstructured":"Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., and Antiga, L. (2019, January 8\u201314). PyTorch: An imperative style, high-performance deep learning library. Proceedings of the Advances in Neural Information Processing Systems, Vancouver, BC, Canada."},{"key":"ref_34","unstructured":"Courville, I.G., Bengio, Y. (2016). Deep Learning, MIT Press."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"92","DOI":"10.1109\/TETCI.2017.2762739","article-title":"Light gated recurrent units for speech recognition","volume":"2","author":"Ravanelli","year":"2018","journal-title":"IEEE Trans. Emerg. Top. Comput. Intell."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"40","DOI":"10.3389\/frai.2020.00040","article-title":"Are GRU cells more specific and LSTM cells more sensitive in motive classification of text?","volume":"3","author":"Gruber","year":"2020","journal-title":"Front. Artif. Intell."},{"key":"ref_37","unstructured":"Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, \u0141., and Polosukhin, I. (2017, January 4\u20139). Attention is all you need. Proceedings of the Advances in Neural Information Processing Systems, Neural Information Processing Systems Foundation, Long Beach, CA, USA."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"229","DOI":"10.1016\/0021-9290(93)90361-H","article-title":"A method for measuring mechanical work and work efficiency during human activities","volume":"26","author":"Sun","year":"1993","journal-title":"J. Biomech."},{"key":"ref_39","unstructured":"Loshchilov, I., and Hutter, F. (2019, January 6\u20139). Decoupled weight decay regularization. Proceedings of the 7th International Conference on Learning Representations, ICLR 2019, OpenReview.net, New Orleans, LA, USA."},{"key":"ref_40","unstructured":"Patterson, J., and Gibson, A. (2017). Deep Learning: A Practitioner\u2019s Approach, O\u2019Reilly."},{"key":"ref_41","doi-asserted-by":"crossref","unstructured":"Akiba, T., Sano, S., Yanase, T., Ohta, T., and Koyama, M. (2019, January 4\u20138). Optuna: A next-generation hyperparameter optimization framework. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Anchorage, AK, USA.","DOI":"10.1145\/3292500.3330701"},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"8272","DOI":"10.1109\/JSEN.2019.2920179","article-title":"Instrumented pyramid adapter for amputee gait analysis and powered prosthesis control","volume":"19","author":"Gabert","year":"2019","journal-title":"IEEE Sens. J."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/23\/18\/7712\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T20:46:23Z","timestamp":1760129183000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/23\/18\/7712"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,9,6]]},"references-count":42,"journal-issue":{"issue":"18","published-online":{"date-parts":[[2023,9]]}},"alternative-id":["s23187712"],"URL":"https:\/\/doi.org\/10.3390\/s23187712","relation":{"has-preprint":[{"id-type":"doi","id":"10.36227\/techrxiv.19790236.v2","asserted-by":"object"},{"id-type":"doi","id":"10.36227\/techrxiv.19790236","asserted-by":"object"},{"id-type":"doi","id":"10.36227\/techrxiv.19790236.v1","asserted-by":"object"}]},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,9,6]]}}}