{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,14]],"date-time":"2026-04-14T01:59:49Z","timestamp":1776131989083,"version":"3.50.1"},"reference-count":38,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2021,11,17]],"date-time":"2021-11-17T00:00:00Z","timestamp":1637107200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Neurosci."],"abstract":"<jats:p>Processing myoelectrical activity in the forearm has for long been considered a promising framework to allow transradial amputees to control motorized prostheses. In spite of expectations, contemporary muscle\u2013computer interfaces built for this purpose typically fail to satisfy one or more important desiderata, such as accuracy, robustness, and\/or naturalness of control, in part due to difficulties in acquiring high-quality signals continuously outside laboratory conditions. In light of such problems, surgically implanted electrodes have been made a viable option that allows for long-term acquisition of intramuscular electromyography (iEMG) measurements of spatially precise origin. As it stands, the question of how information embedded in such signals is best extracted and combined across multiple channels remains open. This study presents and evaluates an approach to this end that uses deep neural networks based on the Long Short-Term Memory (LSTMs) architecture to regress forces exerted by multiple degrees of freedom (DoFs) from multichannel iEMG. Three deep learning models, representing three distinct regression strategies, were evaluated: (I) One-to-One, wherein each DoF is separately estimated by an LSTM model processing a single iEMG channel, (II) All-to-One, wherein each DoF is separately estimated by an LSTM model processing all iEMG channels, and (III) All-to-All, wherein a single LSTM model with access to all iEMG channels estimates all DoFs simultaneously. All models operate on raw iEMG, with no preliminary feature extraction required. When evaluated on a dataset comprising six iEMG channels with concurrent force measurements acquired from 14 subjects, all LSTM strategies were found to significantly outperform a baseline feature-based linear control regression method. This finding indicates that recurrent neural networks can learn to transform raw forearm iEMG signals directly into representations that correlate with forces exerted at the level of the hand to a greater degree than simple features do. Furthermore, the All-to-All and All-to-One strategies were found to exhibit better performance than the One-to-One strategy. This finding suggests that, in spite of the spatially local nature of signals, iEMG from muscles not directly actuating the relevant DoF can provide contextual information that aid in decoding motor intent.<\/jats:p>","DOI":"10.3389\/fnins.2021.777329","type":"journal-article","created":{"date-parts":[[2021,11,17]],"date-time":"2021-11-17T06:55:33Z","timestamp":1637132133000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":10,"title":["End-to-End Estimation of Hand- and Wrist Forces From Raw Intramuscular EMG Signals Using LSTM Networks"],"prefix":"10.3389","volume":"15","author":[{"given":"Alexander E.","family":"Olsson","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Neboj\u0161a","family":"Male\u0161evi\u0107","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Anders","family":"Bj\u00f6rkman","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Christian","family":"Antfolk","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2021,11,17]]},"reference":[{"key":"B1","article-title":"TensorFlow: a system for large-scale machine learning","author":"Abadi","year":"2016","journal-title":"Proceedings of the 12th USENIX Symposium on Operating Systems Design and Implementation, OSDI"},{"key":"B2","doi-asserted-by":"publisher","first-page":"036015","DOI":"10.1088\/1741-2552\/ab0e2e","article-title":"Regression convolutional neural network for improved simultaneous EMG control.","volume":"16","author":"Ameri","year":"2019","journal-title":"J. Neural Eng."},{"key":"B3","doi-asserted-by":"crossref","first-page":"1798","DOI":"10.1109\/TPAMI.2013.50","article-title":"Representation learning: a review and new perspectives.","volume":"35","author":"Bengio","year":"2013","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"B4","doi-asserted-by":"publisher","first-page":"977","DOI":"10.1097\/phm.0b013e3181587f6c","article-title":"Upper-limb prosthetics: critical factors in device abandonment.","volume":"86","author":"Biddiss","year":"2007","journal-title":"Am. J. Phys. Med. Rehabil."},{"key":"B5","first-page":"291","article-title":"Kernel methods","author":"Bishop","year":"2006","journal-title":"Pattern Recognition and Machine Learning"},{"key":"B6","doi-asserted-by":"crossref","first-page":"540","DOI":"10.1080\/10255842.2016.1255943","article-title":"Real-time simulation of hand motion for prosthesis control.","volume":"20","author":"Blana","year":"2017","journal-title":"Comput. Methods Biomech. Biomed. Engin."},{"key":"B7","first-page":"175","article-title":"Osseointegration in skeletal reconstruction and rehabilitation: a review.","volume":"38","author":"Br\u00e5nemark","year":"2001","journal-title":"J. Rehabil. Res. Dev."},{"key":"B8","doi-asserted-by":"publisher","first-page":"549","DOI":"10.1016\/j.neucom.2020.06.139","article-title":"Deep learning for processing electromyographic signals: a taxonomy-based survey.","volume":"452","author":"Buongiorno","year":"2021","journal-title":"Neurocomputing"},{"key":"B9","doi-asserted-by":"publisher","first-page":"828","DOI":"10.1109\/TNSRE.2014.2301234","article-title":"Dexterous control of a prosthetic hand using fine-wire intramuscular electrodes in targeted extrinsic muscles.","volume":"22","author":"Cipriani","year":"2014","journal-title":"IEEE Trans. Neural Syst. Rehabil. Eng."},{"key":"B10","doi-asserted-by":"publisher","first-page":"848","DOI":"10.1109\/TBME.2003.813539","article-title":"A robust, real-time control scheme for multifunction myoelectric control.","volume":"50","author":"Englehart","year":"2003","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"B11","doi-asserted-by":"publisher","first-page":"111","DOI":"10.1109\/tnsre.2007.891391","article-title":"The optimal controller delay for myoelectric prostheses.","volume":"15","author":"Farrell","year":"2007","journal-title":"IEEE Trans. Neural Syst. Rehabil. Eng."},{"key":"B12","doi-asserted-by":"publisher","first-page":"847","DOI":"10.1109\/tbme.2006.889192","article-title":"A comparison of surface and intramuscular myoelectric signal classification.","volume":"54","author":"Hargrove","year":"2007","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"B13","doi-asserted-by":"publisher","first-page":"459","DOI":"10.1054\/jhsb.2001.0587","article-title":"Restoration of tetraplegic hand function by use of the neurocontrol freehand system.","volume":"26","author":"Hobby","year":"2001","journal-title":"J. Hand Surg. Am."},{"key":"B14","doi-asserted-by":"crossref","first-page":"1735","DOI":"10.1162\/neco.1997.9.8.1735","article-title":"Long short-term memory.","volume":"9","author":"Hochreiter","year":"1997","journal-title":"Neural Comput."},{"key":"B15","doi-asserted-by":"crossref","first-page":"82","DOI":"10.1109\/10.204774","article-title":"A new strategy for multifunction myoelectric control.","volume":"40","author":"Hudgins","year":"1993","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"B16","first-page":"150","article-title":"Myoelectric control of artificial limbs: is there a need to change focus? [in the spotlight].","volume":"29","author":"Jiang","year":"2012","journal-title":"IEEE Signal Process. Mag."},{"key":"B17","doi-asserted-by":"publisher","first-page":"143","DOI":"10.1016\/j.jneumeth.2009.09.006","article-title":"Relationship between grasping force and features of single-channel intramuscular EMG signals.","volume":"185","author":"Kamavuako","year":"2009","journal-title":"J. Neurosci. Methods"},{"key":"B18","doi-asserted-by":"publisher","first-page":"619","DOI":"10.1001\/jama.2009.116","article-title":"Targeted muscle reinnervation for real-time myoelectric control of multifunction artificial arms.","volume":"301","author":"Kuiken","year":"2009","journal-title":"JAMA"},{"key":"B19","doi-asserted-by":"publisher","first-page":"58","DOI":"10.3389\/fnbot.2018.00058","article-title":"Causes of performance degradation in non-invasive electromyographic pattern recognition in upper limb prostheses.","volume":"12","author":"Kyranou","year":"2018","journal-title":"Front. Neurorobot."},{"key":"B20","article-title":"Decoupled weight decay regularization","author":"Loshchilov","year":"2019","journal-title":"Proceedings of the 7th International Conference on Learning Representations"},{"key":"B21","doi-asserted-by":"publisher","first-page":"446","DOI":"10.1109\/10.995683","article-title":"A multiple-layer finite-element model of the surface EMG signal IEEE Trans.","volume":"49","author":"Lowery","year":"2002","journal-title":"Biomed. Eng."},{"key":"B22","doi-asserted-by":"publisher","first-page":"1926","DOI":"10.1109\/TBME.2006.881774","article-title":"Simulation of intramuscular EMG signals detected using implantable myoelectric sensors (IMES).","volume":"53","author":"Lowery","year":"2006","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"B23","article-title":"Rectifier nonlinearities improve neural network acoustic models","author":"Maas","year":"2013","journal-title":"Proceedings of the ICML Workshop on Deep Learning for Audio, Speech and Language"},{"key":"B24","doi-asserted-by":"publisher","first-page":"065701","DOI":"10.1088\/1361-6501\/ab0eae","article-title":"Instrumented platform for assessment of isometric hand muscles contractions.","volume":"30","author":"Male\u0161evi\u0107","year":"2019","journal-title":"Meas. Sci. Technol"},{"key":"B25","doi-asserted-by":"publisher","first-page":"10","DOI":"10.1038\/s41597-019-0335-8","article-title":"A database of multi-channel intramuscular electromyogram signals during isometric hand muscles contractions.","volume":"7","author":"Malesevic","year":"2020","journal-title":"Sci. Data"},{"key":"B26","doi-asserted-by":"publisher","first-page":"681","DOI":"10.1109\/TBME.2010.2068298","article-title":"Simultaneous and proportional force estimation for multifunction myoelectric prostheses using mirrored bilateral training.","volume":"58","author":"Nielsen","year":"2011","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"B27","doi-asserted-by":"publisher","DOI":"10.1109\/EMBC.2019.8856648","article-title":"Exploiting the intertemporal structure of the upper-limb semg: comparisons between an LSTM network and cross-sectional myoelectric pattern recognition methods","author":"Olsson","year":"2019","journal-title":"Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS"},{"key":"B28","doi-asserted-by":"publisher","first-page":"287","DOI":"10.1007\/BF02443381","article-title":"Error rate in five-state myoelectric control systems.","volume":"18","author":"Paciga","year":"1980","journal-title":"Med. Biol. Eng. Comput."},{"key":"B29","doi-asserted-by":"crossref","first-page":"21","DOI":"10.3390\/bdcc2030021","article-title":"EMG pattern recognition in the era of big data and deep learning.","volume":"2","author":"Phinyomark","year":"2018","journal-title":"Big Data Cogn. Comput."},{"key":"B30","doi-asserted-by":"publisher","first-page":"969","DOI":"10.3390\/s20040969","article-title":"Deep learning in physiological signal data: a survey.","volume":"20","author":"Rim","year":"2020","journal-title":"Sensors"},{"key":"B31","doi-asserted-by":"crossref","first-page":"255","DOI":"10.3109\/03091902.2016.1167971","article-title":"Recent advancements in prosthetic hand technology.","volume":"40","author":"Saikia","year":"2016","journal-title":"J. Med. Eng. Technol."},{"key":"B32","doi-asserted-by":"publisher","first-page":"643","DOI":"10.1682\/jrrd.2010.09.0177","article-title":"Electromyogram pattern recognition for control of powered upper-limb prostheses: state of the art and challenges for clinical use.","volume":"48","author":"Scheme","year":"2011","journal-title":"J. Rehabil. Res. Dev."},{"key":"B33","doi-asserted-by":"publisher","first-page":"066013","DOI":"10.1088\/1741-2560\/11\/6\/066013","article-title":"Real-time simultaneous and proportional myoelectric control using intramuscular EMG.","volume":"11","author":"Smith","year":"2014","journal-title":"J. Neural Eng."},{"key":"B34","doi-asserted-by":"crossref","first-page":"737","DOI":"10.1109\/TBME.2015.2469741","article-title":"Evaluation of linear regression simultaneous myoelectric control using intramuscular EMG.","volume":"63","author":"Smith","year":"2016","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"B35","first-page":"8","article-title":"Pattern-recognition arm prosthesis: a historical perspective. A final report.","volume":"12","author":"Wirta","year":"1978","journal-title":"Bull. Prosthet. Res"},{"key":"B36","doi-asserted-by":"publisher","first-page":"512","DOI":"10.1109\/jas.2021.1003865","article-title":"Deep learning for EMG-based human-machine interaction: a review.","volume":"8","author":"Xiong","year":"2021","journal-title":"IEEE\/CAA J. Autom. Sin."},{"key":"B37","doi-asserted-by":"crossref","first-page":"459","DOI":"10.1615\/CritRevBiomedEng.v30.i456.80","article-title":"Control of multifunctional prosthetic hands by processing the electromyographic.","volume":"30","author":"Zecca","year":"2017","journal-title":"Signal Crit. Rev. Biomed. Eng."},{"key":"B38","doi-asserted-by":"publisher","first-page":"14105","DOI":"10.1109\/lra.2019.2896518","article-title":"Normalization in training U-Net for 2-D biomedical semantic segmentation.","volume":"11","author":"Zhou","year":"2019","journal-title":"IEEE Robot. Autom. Lett."}],"container-title":["Frontiers in Neuroscience"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnins.2021.777329\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2021,11,17]],"date-time":"2021-11-17T06:55:58Z","timestamp":1637132158000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnins.2021.777329\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,11,17]]},"references-count":38,"alternative-id":["10.3389\/fnins.2021.777329"],"URL":"https:\/\/doi.org\/10.3389\/fnins.2021.777329","relation":{},"ISSN":["1662-453X"],"issn-type":[{"value":"1662-453X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,11,17]]},"article-number":"777329"}}