{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T16:51:27Z","timestamp":1789404687772,"version":"build-2803163510"},"reference-count":30,"publisher":"Public Library of Science (PLoS)","issue":"1","license":[{"start":{"date-parts":[[2026,1,23]],"date-time":"2026-01-23T00:00:00Z","timestamp":1769126400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>Conventional wearable monitoring devices often suffer from insufficient data accuracy and low posture recognition rates, making them inadequate for the demands of professional sports health monitoring. To address these issues, this study proposes a wearable monitoring system for running gait analysis based on the Diffusion Transformer (DiT). The system aims to achieve high-precision running posture recognition and real-time motion monitoring through multi-sensor data fusion and advanced deep learning architecture. First, a wearable system was developed using a nine-axis Micro-Electro-Mechanical System (MEMS) inertial sensor and an UltraWide Band (UWB) positioning module. Data quality was enhanced through sensor calibration, noise compensation, and an adaptive filtering algorithm. Then, a DiT-LSTM running posture recognition model was constructed by integrating the DiT with a Long Short-Term Memory (LSTM) network to perform posture recognition within the wearable system. Experimental results show that on the Human3.6M dataset, the DiT-LSTM model achieved an accuracy of 97.54%, a precision of 97.61%, a recall of 97.73%, an F1-score of 97.58%, and an Area Under the Curve (AUC) of 98.61%. On the HumanEva dataset, the model attained an accuracy of 96.39%, a precision of 96.47%, a recall of 96.8%, an F1-score of 96.92%, and an AUC of 97.9%, all outperforming other algorithms. The complexity assessment showed that DiT-LSTM reached 14.7 GFLOPs on the Human3.6M dataset, with a per-epoch training time of 62.3 seconds, while its per-sample inference latency was only 9.4\u2009ms, meeting real-time monitoring requirements. In the UWB-based position drift correction experiment, Sample 1 achieved mean errors of 0.637 m, 0.581 m, and 0.349 m on the X\/Y\/Z axes, with corresponding RMSE values of 0.041 m, 0.023 m, and 0.025 m, demonstrating high positioning accuracy and stability. By combining multimodal sensors with the DiT-LSTM model, the study offers reliable technical support for running gait analysis, injury prevention, and personalized training guidance.<\/jats:p>","DOI":"10.1371\/journal.pone.0341066","type":"journal-article","created":{"date-parts":[[2026,1,23]],"date-time":"2026-01-23T18:36:26Z","timestamp":1769193386000},"page":"e0341066","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":2,"title":["A wearable monitoring system for running gait analysis by diffusion transformer"],"prefix":"10.1371","volume":"21","author":[{"given":"Xiaoxue","family":"Hu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Guoyu","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-2778-5852","authenticated-orcid":true,"given":"Guodong","family":"Ma","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,1,23]]},"reference":[{"issue":"18","key":"pone.0341066.ref001","doi-asserted-by":"crossref","first-page":"10399","DOI":"10.3390\/app131810399","article-title":"Review on Wearable Technology in Sports: Concepts, Challenges and Opportunities","volume":"13","author":"A\u00c7 Se\u00e7kin","year":"2023","journal-title":"Applied Sciences"},{"key":"pone.0341066.ref002","doi-asserted-by":"crossref","first-page":"1271431","DOI":"10.3389\/fpsyg.2023.1271431","article-title":"Exercise makes better mind: a data mining study on effect of physical activity on academic achievement of college students","volume":"14","author":"S Du","year":"2023","journal-title":"Front Psychol"},{"issue":"2","key":"pone.0341066.ref003","doi-asserted-by":"crossref","first-page":"110","DOI":"10.1111\/sms.14242","article-title":"The load borne by the Achilles tendon during exercise: A systematic review of normative values","volume":"33","author":"Y Demangeot","year":"2023","journal-title":"Scand J Med Sci Sports"},{"issue":"1","key":"pone.0341066.ref004","doi-asserted-by":"crossref","first-page":"57","DOI":"10.1186\/s12984-023-01186-9","article-title":"A systematic review of the applications of markerless motion capture (MMC) technology for clinical measurement in rehabilitation","volume":"20","author":"WWT Lam","year":"2023","journal-title":"J Neuroeng Rehabil"},{"key":"pone.0341066.ref005","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/TIM.2023.3276528","article-title":"Inertial Sensors for Human Motion Analysis: A Comprehensive Review","volume":"72","author":"S Garc\u00eda-de-Villa","year":"2023","journal-title":"IEEE Trans Instrum Meas"},{"issue":"20","key":"pone.0341066.ref006","doi-asserted-by":"crossref","first-page":"6732","DOI":"10.3390\/s24206732","article-title":"Indoor Pedestrian Positioning Method Based on Ultra-Wideband with a Graph Convolutional Network and Visual Fusion","volume":"24","author":"H Mu","year":"2024","journal-title":"Sensors (Basel)"},{"issue":"16","key":"pone.0341066.ref007","doi-asserted-by":"crossref","first-page":"9384","DOI":"10.3390\/app13169384","article-title":"Body-Pose-Guided Action Recognition with Convolutional Long Short-Term Memory (LSTM) in Aerial Videos","volume":"13","author":"SM Saeed","year":"2023","journal-title":"Applied Sciences"},{"key":"pone.0341066.ref008","doi-asserted-by":"crossref","first-page":"164","DOI":"10.1016\/j.neucom.2023.03.001","article-title":"Transformer for Skeleton-based action recognition: A review of recent advances","volume":"537","author":"W Xin","year":"2023","journal-title":"Neurocomputing"},{"issue":"6","key":"pone.0341066.ref009","doi-asserted-by":"crossref","first-page":"1075","DOI":"10.3390\/electronics14061075","article-title":"A Comparison of Deep Learning Techniques for Pose Recognition in Up-and-Go Pole Walking Exercises Using Skeleton Images and Feature Data","volume":"14","author":"W-C Lin","year":"2025","journal-title":"Electronics"},{"issue":"1","key":"pone.0341066.ref010","doi-asserted-by":"crossref","first-page":"40","DOI":"10.3390\/signals4010002","article-title":"Cascading Pose Features with CNN-LSTM for Multiview Human Action Recognition","volume":"4","author":"NR Malik","year":"2023","journal-title":"Signals"},{"issue":"24","key":"pone.0341066.ref011","doi-asserted-by":"crossref","first-page":"19317","DOI":"10.1007\/s00500-023-09321-3","article-title":"An image classification algorithm for football players\u2019 activities using deep neural network","volume":"27","author":"X Li","year":"2023","journal-title":"Soft Comput"},{"issue":"7","key":"pone.0341066.ref012","first-page":"1","article-title":"Attention-based deep learning framework for action recognition in a dark environment","volume":"14","author":"M Munsif","year":"2024","journal-title":"Hum Centric Comput Inf Sci"},{"issue":"12","key":"pone.0341066.ref013","doi-asserted-by":"crossref","first-page":"36975","DOI":"10.1007\/s11042-023-16388-5","article-title":"A visual intelligent system for students\u2019 behavior classification using body pose and facial features in a smart classroom","volume":"83","author":"C Pabba","year":"2023","journal-title":"Multimed Tools Appl"},{"issue":"2","key":"pone.0341066.ref014","doi-asserted-by":"crossref","first-page":"15","DOI":"10.3390\/technologies12020015","article-title":"A Review of Machine Learning and Deep Learning for Object Detection, Semantic Segmentation, and Human Action Recognition in Machine and Robotic Vision","volume":"12","author":"N Manakitsa","year":"2024","journal-title":"Technologies"},{"issue":"5","key":"pone.0341066.ref015","doi-asserted-by":"crossref","first-page":"3841","DOI":"10.1007\/s10489-024-05325-0","article-title":"Pose pattern mining using transformer for motion classification","volume":"54","author":"S-E Lee","year":"2024","journal-title":"Appl Intell"},{"issue":"7","key":"pone.0341066.ref016","doi-asserted-by":"crossref","DOI":"10.1061\/JCEMD4.COENG-13638","article-title":"Kinesiology-Inspired Assessment of Intrusion Risk Based on Human Motion Features","volume":"150","author":"H Huang","year":"2024","journal-title":"J Constr Eng Manage"},{"issue":"2","key":"pone.0341066.ref017","doi-asserted-by":"crossref","first-page":"106","DOI":"10.15561\/20755279.2024.0208","article-title":"Applied biomechanics within the Kinesiology discipline in higher education","volume":"28","author":"V Potop","year":"2024","journal-title":"Physical Education of Students"},{"key":"pone.0341066.ref018","doi-asserted-by":"crossref","first-page":"115193","DOI":"10.1016\/j.sna.2024.115193","article-title":"Electromechanical performance of flexible sensor based on 3D printed biomimetic all-PDMS-coated multistage structure: Toward kinesiology monitoring and human-machine interaction","volume":"369","author":"X Guo","year":"2024","journal-title":"Sensors and Actuators A: Physical"},{"key":"pone.0341066.ref019","doi-asserted-by":"crossref","first-page":"1335272","DOI":"10.3389\/fspor.2024.1335272","article-title":"Assessing the validity of two-dimensional video analysis for measuring lower limb joint angles during fencing lunge","volume":"6","author":"K Chida","year":"2024","journal-title":"Front Sports Act Living"},{"issue":"4","key":"pone.0341066.ref020","doi-asserted-by":"crossref","first-page":"148","DOI":"10.3390\/technologies13040148","article-title":"Autonomous Real-Time Mass Center Location and Inertia Identification for Grappling Space Robotics","volume":"13","author":"T Sands","year":"2025","journal-title":"Technologies"},{"key":"pone.0341066.ref021","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1016\/j.gaitpost.2024.04.006","article-title":"A novel method for accurate division of the gait cycle into seven phases using shank angular velocity","volume":"111","author":"M Salminen","year":"2024","journal-title":"Gait Posture"},{"issue":"1","key":"pone.0341066.ref022","doi-asserted-by":"crossref","first-page":"114001","DOI":"10.1016\/j.measurement.2023.114001","article-title":"A systematic review of user-conducted calibration methods for MEMS-based IMUs","volume":"225","author":"A Harindranath","year":"2024","journal-title":"Measurement"},{"issue":"4","key":"pone.0341066.ref023","doi-asserted-by":"crossref","first-page":"4851","DOI":"10.1109\/JSEN.2023.3346400","article-title":"Data Snooping for Pose Estimation Based on Generalized Errors-in-Variables Model","volume":"24","author":"C Zhang","year":"2024","journal-title":"IEEE Sensors J"},{"issue":"3","key":"pone.0341066.ref024","doi-asserted-by":"crossref","first-page":"346","DOI":"10.3390\/mi15030346","article-title":"Attitude Algorithm of Gyroscope-Free Strapdown Inertial Navigation System Using Kalman Filter","volume":"15","author":"X Jiang","year":"2024","journal-title":"Micromachines (Basel)"},{"issue":"7","key":"pone.0341066.ref025","doi-asserted-by":"crossref","first-page":"1678","DOI":"10.3390\/electronics12071678","article-title":"Survey on NLOS Identification and Error Mitigation for UWB Indoor Positioning","volume":"12","author":"F Wang","year":"2023","journal-title":"Electronics"},{"issue":"1","key":"pone.0341066.ref026","doi-asserted-by":"crossref","first-page":"73","DOI":"10.1007\/s11044-022-09855-8","article-title":"Optimal design of active-passive shoulder exoskeletons: a computational modeling of human-robot interaction","volume":"57","author":"A Nasr","year":"2022","journal-title":"Multibody Syst Dyn"},{"key":"pone.0341066.ref027","article-title":"State of charge estimation of lithium-ion battery based on extended Kalman filter algorithm","volume":"11","author":"J Xie","year":"2023","journal-title":"Front Energy Res"},{"issue":"5","key":"pone.0341066.ref028","doi-asserted-by":"crossref","first-page":"747","DOI":"10.3390\/jmse12050747","article-title":"A Robust Vector-Tracking Loop Based on KF and RTS Smoothing for Shipborne Navigation","volume":"12","author":"Y Hu","year":"2024","journal-title":"Journal of Marine Science and Engineering"},{"issue":"23","key":"pone.0341066.ref029","doi-asserted-by":"crossref","first-page":"26417","DOI":"10.1007\/s11042-024-20179-x","article-title":"Exploiting multi-transformer encoder with multiple-hypothesis aggregation via diffusion model for 3D human pose estimation","volume":"84","author":"S Arthanari","year":"2024","journal-title":"Multimed Tools Appl"},{"issue":"11","key":"pone.0341066.ref030","doi-asserted-by":"crossref","first-page":"33019","DOI":"10.1007\/s11042-023-17036-8","article-title":"Real-time yoga pose classification with 3-D pose estimation model with LSTM","volume":"83","author":"RP Srivastava","year":"2023","journal-title":"Multimed Tools Appl"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0341066","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,23]],"date-time":"2026-01-23T18:36:31Z","timestamp":1769193391000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0341066"}},"subtitle":[],"editor":[{"given":"Anwar","family":"P.P. Abdul Majeed","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,1,23]]},"references-count":30,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,1,23]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0341066","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,1,23]]}}}