{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T23:11:27Z","timestamp":1791501087968,"version":"4.3.4"},"reference-count":55,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,11,7]],"date-time":"2025-11-07T00:00:00Z","timestamp":1762473600000},"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. Bioeng. Biotechnol."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Advancements in artificial intelligence are transforming rehabilitation by enabling scalable, patient-centric solutions within modern healthcare systems. This study introduces 3D-PoseFormer, a deep multimodal framework designed for the telerehabilitation of individuals with lower back pain (LBP).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>The proposed system performs automated data acquisition using synchronized RGB and depth video streams to enable real-time, markerless, and sensor-free analysis of physiotherapy exercises. From the depth sensing module, 3D body joint positions are extracted and used to generate SMPL-based mesh vertices for detailed biomechanical analysis and postural representation. Simultaneously, RGB frames are processed using keypoint detection algorithms\u2014Shi-Tomasi, AKAZE, BRISK, SIFT, and Harris corner detection. Extracted features are enhanced through semantic contour analysis of segmented body parts to capture localized appearance-based information relevant to LBP therapy. The fused multimodal features are then passed to a Transformer-based machine learning model that captures temporal motion patterns for accurate exercise classification and human intention recognition.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The system removes the need for wearable sensors and supports autonomous, continuous monitoring in home-based rehabilitation. Validation on the KIMORE dataset (baseline, including rehabilitation exercises by patients with lower back pain), mRI dataset (rehabilitation exercises), and UTKinect-Action3D dataset (comprising diverse subjects and activity scenarios) achieved state-of-the-art accuracies of 94.73%, 91%, and 94.2%, respectively.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>Results demonstrate the robustness, generalizability, and clinical potential of 3D-PoseFormer in AI-assisted rehabilitation, offering a scalable and intelligent healthcare system for remote physiotherapy and patient monitoring.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fbioe.2025.1631910","type":"journal-article","created":{"date-parts":[[2025,11,7]],"date-time":"2025-11-07T06:22:28Z","timestamp":1762496548000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["Deep multimodal biomechanical analysis for lower back pain rehabilitation to improve patients stability"],"prefix":"10.3389","volume":"13","author":[{"given":"Muhammad Abrar","family":"Ashraf","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yanfeng","family":"Wu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shaheryar","family":"Najam","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mohammed","family":"Alshehri","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yahya","family":"AlQahtani","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hanan","family":"Aljuaid","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ahmad","family":"Jalal","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hui","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,11,7]]},"reference":[{"key":"B65","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/ACCESS.2023.3239692","article-title":"Body-worn sensors for recognizing physical sports activities in exergaming via deep learning model","volume":"2023","author":"Afsar","year":"2023","journal-title":"IEEE Access"},{"key":"B1","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/C-CODE58145.2023.10139863","article-title":"Deep skeleton modeling and hybrid hand-crafted cues over physical exercises","volume":"2023","author":"Akhter","year":"2023","journal-title":"Proc. Int. Conf. Commun. Comput. Digital Syst. (C-CODE)"},{"key":"B2","doi-asserted-by":"publisher","first-page":"13.1","DOI":"10.5244\/C.27.13","article-title":"Fast explicit diffusion for accelerated features in nonlinear scale spaces","volume":"2011","author":"Alcantarilla","year":"2011","journal-title":"Proc. Br. Mach. Vis. Conf. (BMVC)"},{"key":"B4","volume-title":"mRI: multi-Modal 3D human pose estimation dataset using mmWave, RGB-D, and inertial sensors","author":"An","year":"2022"},{"key":"B5","doi-asserted-by":"publisher","first-page":"495","DOI":"10.20517\/ir.2023.28","article-title":"Muscle synergy analysis for gesture recognition based on sEMG images and shapley value","volume":"3","author":"Ao","year":"2023","journal-title":"Intell. Robot."},{"key":"B6","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/ACCESS.2025.3526710","article-title":"A novel telerehabilitation system for physical exercise monitoring in elderly healthcare","volume":"2025","author":"Ashraf","year":"2025","journal-title":"IEEE Access"},{"key":"B61","first-page":"02034","article-title":"Action recognition based on 2D skeletons extracted from RGB videos","volume-title":"Proc. MATEC Web Conf","author":"Aubry","year":"2019"},{"key":"B7","doi-asserted-by":"publisher","first-page":"33","DOI":"10.1109\/ICET63392.2024.10935118","article-title":"Robust exercise-based telerehabilitation for elderly healthcare services","volume":"2024","author":"Awan","year":"2024","journal-title":"Proc. Int. Conf. Emerg. Trends Electr. Control, Telecommun. Eng."},{"key":"B56","first-page":"199","article-title":"Kinect-based platform for movement monitoring and fall-detection of elderly people","author":"Barabas","year":"2019"},{"key":"B9","doi-asserted-by":"publisher","first-page":"7291","DOI":"10.1109\/CVPR.2017.143","article-title":"Realtime multi-person 2D pose estimation using part affinity fields","author":"Cao","year":"2017","journal-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit."},{"key":"B57","doi-asserted-by":"publisher","first-page":"1436","DOI":"10.1109\/TNSRE.2019.2923060","article-title":"The kimore dataset: kinematic assessment of movement and clinical scores for remote monitoring of physical rehabilitation","volume":"27","author":"Capecci","year":"2019","journal-title":"IEEE Trans. Neural Syst. Rehabil. Eng."},{"key":"B10","doi-asserted-by":"publisher","first-page":"2566","DOI":"10.1016\/j.ridd.2011.07.002","article-title":"A kinect-based system for physical rehabilitation: a pilot study for young adults with motor disabilities","volume":"32","author":"Chang","year":"2011","journal-title":"Res. Dev. Disabil."},{"key":"B12","doi-asserted-by":"publisher","first-page":"110","DOI":"10.1049\/iet-cvi.2017.0031","article-title":"Human action recognition using similarity degree between postures and spectral learning","volume":"12","author":"Ding","year":"2018","journal-title":"IET Comput. Vis."},{"key":"B13","doi-asserted-by":"publisher","first-page":"112","DOI":"10.3138\/fm57-6770-u75u-7727","article-title":"Algorithms for the reduction of the number of points required to represent a digitized line or its caricature","volume":"10","author":"Douglas","year":"1973","journal-title":"Can. Cartogr."},{"key":"B14","first-page":"1","article-title":"A novel full-body and geometric features for physical sports interaction recognition","volume-title":"Proc. International conference on innovative computing","author":"Fatima","year":"2024"},{"key":"B64","doi-asserted-by":"publisher","first-page":"99152","DOI":"10.1109\/ACCESS.2019.2927134","article-title":"A hybrid deep learning model for human activity recognition using multimodal body sensing data","volume":"7","author":"Gumaei","year":"2019","journal-title":"IEEE Access"},{"key":"B16","first-page":"43","article-title":"Human activity recognition using RGB video: a survey","volume-title":"Proc. Int. Conf. Soft Computing and Pattern Recognition (SoCPaR)","author":"Gupta","year":"2020"},{"key":"B17","first-page":"181","article-title":"DeepLabv3","author":"Hamamoto","year":"2024"},{"key":"B59","doi-asserted-by":"crossref","first-page":"366","DOI":"10.1109\/IMSA61967.2024.10652817","article-title":"Enhancing physical therapy through transformer-based models: a study on exercise classification","volume-title":"2024 Intelligent Methods, Systems, and Applications (IMSA)","author":"Hamdy","year":"2024"},{"key":"B18","doi-asserted-by":"publisher","first-page":"147","DOI":"10.5244\/C.2.23","article-title":"A combined corner and edge detector","volume":"1988","author":"Harris","year":"1988","journal-title":"Proc. Alvey Vis. Conf."},{"key":"B19","volume-title":"Single-Human-Parsing-LIP: body part labeling for human parsing","author":"Huang","year":"2024"},{"key":"B63","first-page":"1","article-title":"Automated Gestures Recognition in Exergaming","volume-title":"Proc. 2022 Int. Conf","author":"Javeed","year":"2022"},{"key":"B20","doi-asserted-by":"publisher","DOI":"10.57197\/jdr-2023-0065","article-title":"Artificial intelligence-driven remote monitoring model for physical rehabilitation","volume":"3","author":"Jleli","year":"2024","journal-title":"J. Disabil. Res."},{"key":"B21","doi-asserted-by":"publisher","first-page":"81279","DOI":"10.1109\/ACCESS.2025.3565024","article-title":"Holistic pose estimation and dynamic motion analysis for telerehabilitation of physically disabled individuals","volume":"2025","author":"Kamal","year":"2025","journal-title":"IEEE Access"},{"key":"B22","doi-asserted-by":"publisher","first-page":"7122","DOI":"10.1109\/cvpr.2018.00744","article-title":"End-to-End recovery of human shape and pose","volume":"2018","author":"Kanazawa","year":"2018","journal-title":"Proc. CVPR"},{"key":"B66","doi-asserted-by":"publisher","first-page":"2","DOI":"10.1109\/ComTech65062.2025.11034496","article-title":"Wearable sensors for exergaming physical exercise monitoring via dynamic features","volume":"2025","author":"Kaynat","year":"2025","journal-title":"ComTech"},{"key":"B23","doi-asserted-by":"publisher","first-page":"3567","DOI":"10.3390\/electronics11213567","article-title":"3D skeletal volume templates for deep learning-based activity recognition","volume":"11","author":"Ke\u00e7eli","year":"2022","journal-title":"Electronics"},{"key":"B67","doi-asserted-by":"publisher","first-page":"2","DOI":"10.1109\/ComTech65062.2025.11034569","article-title":"Health gaming based activity recognition using body-worn sensors via artificial neural network","volume":"2025","author":"Khan","year":"2025","journal-title":"ComTech"},{"key":"B24","doi-asserted-by":"publisher","first-page":"5253","DOI":"10.48550\/arXiv.1912.05656","article-title":"VIBE: video inference for human body pose and shape estimation","volume":"2020","author":"Kocabas","year":"2020","journal-title":"Proc. CVPR"},{"key":"B25","doi-asserted-by":"crossref","DOI":"10.1063\/5.0111612","article-title":"Human action recognition from depth sensor via skeletal joint and shape trajectories with a time-series graph matching","author":"Kumar","year":"2024"},{"key":"B26","first-page":"2548","article-title":"BRISK: binary robust invariant scalable keypoints","author":"Leutenegger","year":"2011"},{"key":"B27","first-page":"2354","article-title":"Action recognition from RGB video using pose estimation and graph CNNs","volume-title":"Proc. IEEE Int. Conf. Image Processing (ICIP)","author":"Li","year":"2021"},{"key":"B28","doi-asserted-by":"publisher","first-page":"3202","DOI":"10.48550\/arXiv.2106.13230","article-title":"Video swin transformer","volume":"2022","author":"Liu","year":"2022","journal-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit."},{"key":"B29","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/2816795.2818013","article-title":"SMPL: a skinned multi-person linear model","volume":"34","author":"Loper","year":"2015","journal-title":"ACM Trans. Graph."},{"key":"B30","doi-asserted-by":"publisher","first-page":"91","DOI":"10.1023\/b:visi.0000029664.99615.94","article-title":"Distinctive image features from scale-invariant keypoints","volume":"60","author":"Lowe","year":"2004","journal-title":"Int. J. Comput. Vis."},{"key":"B68","doi-asserted-by":"publisher","first-page":"115","DOI":"10.3390\/sym12111766","article-title":"Accurate physical activity recognition using multidimensional features and markov model for smart health fitness","volume":"12","author":"Nadeem","year":"2020","journal-title":"Symmetry"},{"key":"B33","first-page":"6","article-title":"Wearable sensors-based activity classification for intelligent healthcare monitoring","volume-title":"Proc. International conference on advancements in computational sciences (ICACS)","author":"Nazar","year":"2025"},{"key":"B34","doi-asserted-by":"publisher","first-page":"105021","DOI":"10.1016\/j.imavis.2024.105021","article-title":"A ResNet-101 deep learning framework induced transfer learning strategy for moving object detection","volume":"146","author":"Panigrahi","year":"2024","journal-title":"Image Vis. Comput."},{"key":"B35","doi-asserted-by":"publisher","first-page":"7025","DOI":"10.48550\/arXiv.1611.07828","article-title":"Coarse-to-Fine volumetric prediction for single-image 3D human pose","volume":"2017","author":"Pavlakos","year":"2017","journal-title":"Proc. CVPR"},{"key":"B37","doi-asserted-by":"publisher","first-page":"2564","DOI":"10.1109\/iccv.2011.6126544","article-title":"ORB: an efficient alternative to SIFT or SURF","volume":"2011","author":"Rublee","year":"2011","journal-title":"Proc. IEEE Int. Conf. Comput. Vis. (ICCV)"},{"key":"B3","doi-asserted-by":"publisher","first-page":"91","DOI":"10.14744\/bauh.2023.43531","article-title":"Evaluation of wearable technologies in physiotherapy and rehabilitation","author":"\u015eahin","year":"2024","journal-title":"BAU Health Innov."},{"key":"B62","doi-asserted-by":"publisher","first-page":"1464","DOI":"10.3390\/s22041464","article-title":"3DMesh-GAR: 3D human body mesh-based method for group activity recognition","volume":"22","author":"Saqlain","year":"2022","journal-title":"Sensors"},{"key":"B39","first-page":"593","article-title":"Good features to track","volume-title":"Proc. IEEE conf. Computer vision and pattern recognition (CVPR)","author":"Shi","year":"1994"},{"key":"B41","first-page":"9","article-title":"Disabled rehabilitation monitoring and patients healthcare recognition using machine learning","volume-title":"Proc. International conference on advancements in computational sciences (ICACS)","author":"Tayyab","year":"2025"},{"key":"B42","doi-asserted-by":"publisher","first-page":"441","DOI":"10.3390\/s25020441","article-title":"A hybrid approach for sports activity recognition using key body descriptors and hybrid deep learning classifier","volume":"25","author":"Tayyab","year":"2025","journal-title":"Sensors"},{"key":"B43","doi-asserted-by":"publisher","first-page":"5998","DOI":"10.48550\/arXiv.1706.03762","article-title":"Attention is all you need","volume":"2017","author":"Vaswani","year":"2017","journal-title":"Proc. Adv. Neural Inf. Process. Syst."},{"key":"B44","doi-asserted-by":"publisher","first-page":"3169","DOI":"10.1109\/CVPR.2011.5995407","article-title":"Action recognition by dense trajectories","volume":"2011","author":"Wang","year":"2011","journal-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit."},{"key":"B46","doi-asserted-by":"publisher","first-page":"101753","DOI":"10.1111\/aor.13004","article-title":"EMG-based estimation of limb movement using deep learning with recurrent convolutional neural networks","volume":"102","author":"Xia","year":"2020","journal-title":"Artif. Intell. Med."},{"key":"B60","doi-asserted-by":"publisher","first-page":"136","DOI":"10.1016\/j.patrec.2021.06.003","article-title":"Scene image and human skeleton-based dual-stream human action recognition","volume":"148","author":"Xu","year":"2021","journal-title":"Pattern Recognit. Lett."},{"key":"B47","doi-asserted-by":"publisher","first-page":"1057","DOI":"10.1145\/2393347.2396382","article-title":"Recognizing actions using depth motion maps-based histograms of oriented gradients","volume":"2012","author":"Yang","year":"2012","journal-title":"ACM Int. Conf. Multimed."},{"key":"B49","doi-asserted-by":"publisher","first-page":"1261","DOI":"10.1007\/s11042-024-19092-0","article-title":"Unlocking the potential of RNN and CNN models for accurate rehabilitation exercise classification on multi-datasets","volume":"84","author":"Zaher","year":"2024","journal-title":"Multimed. Tools Appl."},{"key":"B50","doi-asserted-by":"publisher","first-page":"382","DOI":"10.1080\/24751839.2025.2454053","article-title":"Rehabilitation monitoring and assessment: a comparative analysis of feature engineering and machine learning algorithms on the UI-PRMD and KIMORE benchmark datasets","volume":"2025","author":"Zaher","year":"2025","journal-title":"J. Inf. Telecommun."},{"key":"B52","doi-asserted-by":"publisher","first-page":"2148","DOI":"10.1109\/cvpr.2018.00229","article-title":"Monocular 3D pose and shape estimation of multiple people in natural scenes\u2014the importance of multiple scene constraints","volume":"2018","author":"Zanfir","year":"2018","journal-title":"Proc. CVPR"},{"key":"B53","doi-asserted-by":"publisher","first-page":"236","DOI":"10.1145\/357994.358023","article-title":"A fast parallel algorithm for thinning digital patterns","volume":"27","author":"Zhang","year":"1984","journal-title":"Commun. ACM"},{"key":"B54","first-page":"100104","article-title":"Physical activity recognition for elderly using hybrid deep models with single IMU","volume":"7","author":"Zhang","year":"2019","journal-title":"IEEE Access"}],"container-title":["Frontiers in Bioengineering and Biotechnology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbioe.2025.1631910\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,7]],"date-time":"2025-11-07T06:22:32Z","timestamp":1762496552000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbioe.2025.1631910\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,7]]},"references-count":55,"alternative-id":["10.3389\/fbioe.2025.1631910"],"URL":"https:\/\/doi.org\/10.3389\/fbioe.2025.1631910","relation":{},"ISSN":["2296-4185"],"issn-type":[{"value":"2296-4185","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11,7]]},"article-number":"1631910"}}