{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,10]],"date-time":"2026-06-10T21:59:31Z","timestamp":1781128771576,"version":"3.54.1"},"reference-count":58,"publisher":"MDPI AG","issue":"3","license":[{"start":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T00:00:00Z","timestamp":1772064000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Bioengineering"],"abstract":"<jats:p>Pediatric gait abnormalities are closely intertwined with musculoskeletal dysfunctions and heightened injury risk, underscoring the urgency of early and accessible screening tools. Here, we develop and validate a video-based semi-supervised Abnormal Gait Recognition Module (AGRM) to address unmet needs in pediatric gait assessment, with a focus on diagnostic performance and clinical interpretability. The AGRM is built on a 3D ResNet backbone, synergistically integrated with a Mean Teacher Module (MTM) to mitigate the limitations of limited labeled clinical data, and a Spatial Hierarchical Pooling Module (SHPM) for robust multiscale spatiotemporal feature extraction\u2014two core innovations tailored to gait dynamics. We trained and validated the model on a hybrid dataset combining self-collected pediatric gait videos and the public CASIA-B dataset, evaluating its performance in binary (normal vs. abnormal) and three-class (normal, genu varum, genu valgum) classification tasks using accuracy, macro-precision, macro-recall, and macro-F1 score. Ablation studies quantified the incremental contributions of MTM and SHPM, while Grad-CAM visualization was employed to enhance model interpretability. In the three-class classification task, the AGRM achieved a 70.5% accuracy, 72.1% macro-precision, 71.5% macro-recall, and a macro-F1 score of 0.718; in the binary task, it yielded a 80.3% precision and 79.2% recall. SHPM significantly augmented spatiotemporal feature aggregation, capturing fine-grained gait dynamics, whereas MTM improved model generalization under constrained labeled data scenarios\u2014findings corroborated by ablation experiments. Grad-CAM visualization confirmed the model\u2019s targeted attention to lower extremity regions, particularly the knee joints, aligning with the pathological loci of gait abnormalities. Collectively, our AGRM demonstrates robust performance and generalization in identifying pediatric gait abnormalities, while effectively capturing key pathological gait characteristics. This video-based intelligent approach offers a promising tool for early gait screening in both clinical and community settings, addressing barriers to accessible pediatric musculoskeletal assessment.<\/jats:p>","DOI":"10.3390\/bioengineering13030272","type":"journal-article","created":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T16:57:13Z","timestamp":1772125033000},"page":"272","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":2,"title":["Development and Performance Analysis of a Semi-Supervised Gait Recognition Model for Pediatric Abnormalities Using a Hybrid Dataset"],"prefix":"10.3390","volume":"13","author":[{"given":"Xiaoneng","family":"Song","sequence":"first","affiliation":[{"name":"Department of Physical Education, Jiangnan University, Wuxi 214122, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2702-9500","authenticated-orcid":false,"given":"Kun","family":"Qian","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence and Computer Science, Jiangnan University, Wuxi 214122, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3626-2606","authenticated-orcid":false,"given":"Sida","family":"Tang","sequence":"additional","affiliation":[{"name":"School of Electronic and Information Engineering, Wuxi University, Wuxi 214105, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,2,26]]},"reference":[{"key":"ref_1","first-page":"1333","article-title":"General methods and development actuality of gait recognition","volume":"Volume 3","author":"Li","year":"2007","journal-title":"2007 International Conference on Wavelet Analysis and Pattern Recognition"},{"key":"ref_2","unstructured":"Lehnert\u2013Schroth, C., and Auner\u2013Gr\u00f6bl, P. (2021). Dreidimensionale Skoliosebehandlung: Atmungs\u2013Orthop\u00e4die System Schroth, Elsevier Health Sciences."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"1204","DOI":"10.1002\/art.22515","article-title":"Association between valgus and varus alignment and the development and progression of radio- graphic osteoarthritis of the knee","volume":"56","author":"Brouwer","year":"2007","journal-title":"Arthritis Rheum."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"1002","DOI":"10.1002\/art.30216","article-title":"Varus\u2013valgus alignment: Reduced risk of subsequent cartilage loss in the less loaded compartment","volume":"63","author":"Moisio","year":"2011","journal-title":"Arthritis Rheum."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"188","DOI":"10.1001\/jama.286.2.188","article-title":"The role of knee alignment in disease progression and functional decline in knee osteoarthritis","volume":"286","author":"Sharma","year":"2001","journal-title":"JAMA"},{"key":"ref_6","first-page":"e2020121","article-title":"The prevalence of genu varum and genu valgum in overweight and obese patients: Assessing the relationship between body mass index and knee angular deformities","volume":"91","author":"Soheilipour","year":"2020","journal-title":"Acta Biomed. Atenei Parm."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"544","DOI":"10.1016\/j.knee.2013.09.008","article-title":"Factors contributing to inherent varus alignment of lower limb in normal Asian adults: Role of tibial plateau inclination","volume":"21","author":"Shetty","year":"2014","journal-title":"Knee"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"631","DOI":"10.1016\/j.gaitpost.2013.02.011","article-title":"Intra-individual gait speed variability in healthy children aged 1\u201315 years","volume":"38","author":"Baur","year":"2013","journal-title":"Gait Posture"},{"key":"ref_9","first-page":"13","article-title":"Gait recognition methods in the task of biometric human identification","volume":"6","author":"Purish","year":"2023","journal-title":"Her. Adv. Inf. Technol."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"2335","DOI":"10.2147\/NDT.S266585","article-title":"Can quantitative gait analysis be used to guide treatment of patients with different subtypes of Parkinson\u2019s disease?","volume":"16","author":"Wu","year":"2020","journal-title":"Neuropsychiatr. Dis. Treat."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"165","DOI":"10.3390\/biomechanics4010011","article-title":"Gait biomechanical parameters related to falls in the elderly: A systematic review","volume":"4","author":"Silva","year":"2024","journal-title":"Biomechanics"},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3230633","article-title":"A survey on gait recognition","volume":"51","author":"Wan","year":"2018","journal-title":"ACM Comput. Surv."},{"key":"ref_13","first-page":"264","article-title":"Deep gait recognition: A survey","volume":"45","author":"Etemad","year":"2022","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"1950027","DOI":"10.1142\/S0129065719500278","article-title":"Human gait recognition based on frame-by-frame gait energy images and convolutional long short\u2013term memory","volume":"30","author":"Wang","year":"2020","journal-title":"Int. J. Neural Syst."},{"key":"ref_15","unstructured":"Li, K., Wang, Y., Gao, P., Song, G., Liu, Y., Li, H., and Qiao, Y. (2022). UniFormer: Unified transformer for efficient spatiotemporal representation learning. arXiv."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Tran, D., Bourdev, L., Fergus, R., Torresani, L., and Paluri, M. (2015). Learning spatiotemporal features with 3D convolutional networks. Proceedings of the IEEE International Conference on Computer Vision, IEEE.","DOI":"10.1109\/ICCV.2015.510"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"12581","DOI":"10.1109\/TPAMI.2023.3282631","article-title":"UniFormer: Unifying convolution and self-attention for visual recognition","volume":"45","author":"Li","year":"2023","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Yang, Q., Wei, X., Wang, B., Hua, X.S., and Zhang, L. (2021). Interactive self-training with mean teachers for semi-supervised object detection. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, IEEE.","DOI":"10.1109\/CVPR46437.2021.00588"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"7057","DOI":"10.1007\/s00521-024-09445-z","article-title":"GaitASMS: Gait recognition by adaptive structured spatial representation and multi\u2013scale temporal aggregation","volume":"36","author":"Sun","year":"2024","journal-title":"Neural Comput. Appl."},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Wang, L., Liu, B., Wang, B., and Yu, F. (2023). GAITMM: Multi-granularity motion sequence learning for gait recognition. 2023 IEEE International Conference on Image Processing (ICIP), IEEE.","DOI":"10.1109\/ICIP49359.2023.10221893"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"107622","DOI":"10.1016\/j.patcog.2020.107622","article-title":"Cascaded hierarchical atrous spatial pyramid pooling module for semantic segmentation","volume":"110","author":"Lian","year":"2021","journal-title":"Pattern Recognit."},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Pin\u010di\u0107, D., Su\u0161anj, D., and Lenac, K. (2022). Gait recognition with self-supervised learning of gait features based on vision transformers. Sensors, 22.","DOI":"10.3390\/s22197140"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"149","DOI":"10.1016\/j.gaitpost.2011.03.027","article-title":"Efficacy of clinical gait analysis: A systematic review","volume":"34","author":"Wren","year":"2011","journal-title":"Gait Posture"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"81","DOI":"10.1007\/s00508-016-1096-4","article-title":"Gait disorders in adults and the elderly: A clinical guide","volume":"129","author":"Pirker","year":"2017","journal-title":"Wien. Klin. Wochenschr."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"618","DOI":"10.1111\/dmcn.13722","article-title":"The Gait Outcomes Assessment List (GOAL): Validation of a new assessment of gait function for children with cerebral palsy","volume":"60","author":"Thomason","year":"2018","journal-title":"Dev. Med. Child Neurol."},{"key":"ref_26","first-page":"1000","article-title":"Efficient night gait recognition based on template matching","volume":"Volume 3","author":"Tan","year":"2006","journal-title":"18th International Conference on Pattern Recognition (ICPR)"},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"4891","DOI":"10.1609\/aaai.v38i5.28292","article-title":"Semi\u2013supervised active learning for video action detection","volume":"Volume 38","author":"Singh","year":"2024","journal-title":"Proceedings of the AAAI Conference on Artificial Intelligence"},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Zhao, B., Cui, Q., Song, R., Qiu, Y., and Liang, J. (2022). Decoupled knowledge distillation. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, IEEE.","DOI":"10.1109\/CVPR52688.2022.01165"},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"1849","DOI":"10.1007\/s11548-023-02886-2","article-title":"Deep learning\u2013based overall survival prediction model in patients with rare cancer: A case study for primary central nervous system lymphoma","volume":"18","author":"She","year":"2023","journal-title":"Int. J. Comput. Assist. Radiol. Surg."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"8295","DOI":"10.1609\/aaai.v33i01.33018295","article-title":"Horizontal pyramid matching for person re-identification","volume":"Volume 33","author":"Fu","year":"2019","journal-title":"Proceedings of the AAAI Conference on Artificial Intelligence"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Dave, I.R., Rizve, M.N., Chen, C., and Shah, M. (2023). TimeBalance: Temporally-invariant and temporally-distinctive video representations for semi-supervised action recognition. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, IEEE.","DOI":"10.1109\/CVPR52729.2023.00232"},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"291","DOI":"10.1007\/s11263-024-02186-5","article-title":"Learning rate curriculum","volume":"133","author":"Croitoru","year":"2025","journal-title":"Int. J. Comput. Vis."},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Bengio, Y. (2012). Practical recommendations for gradient\u2013based training of deep architectures. Neural Networks: Tricks of the Trade, Springer. [2nd ed.].","DOI":"10.1007\/978-3-642-35289-8_26"},{"key":"ref_34","unstructured":"Kynk\u00e4\u00e4nniemi, T., Karras, T., Laine, S., Lehtinen, J., and Aila, T. (2019). Improved precision and recall metric for assessing generative models. Advances in Neural Information Processing Systems, NeurIPS Foundation."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"168","DOI":"10.1016\/j.aci.2018.08.003","article-title":"Classification assessment methods","volume":"17","author":"Tharwat","year":"2021","journal-title":"Appl. Comput. Inform."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"7","DOI":"10.1016\/j.clinbiomech.2015.10.008","article-title":"Three-dimensional hip and knee kinematics during walking, running, and single-limb drop landing in females with and without genu valgum","volume":"31","author":"Barrios","year":"2016","journal-title":"Clin. Biomech."},{"key":"ref_37","first-page":"3168","article-title":"Kinetic and kinematic analysis of gait pattern of 13-year-old children with unilateral genu valgum","volume":"20","author":"Ganesan","year":"2016","journal-title":"Eur. Rev. Med. Pharmacol. Sci."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"522","DOI":"10.1016\/j.joca.2007.08.008","article-title":"Knee varus\u2013valgus motion during gait\u2014a measure of joint stability in patients with osteoarthritis?","volume":"16","author":"Steultjens","year":"2008","journal-title":"Osteoarthr. Cartil."},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"1191","DOI":"10.1007\/s00264-014-2312-3","article-title":"Gait knee kinematic alterations in medial osteoarthritis: Three dimensional assessment","volume":"38","author":"Bytyqi","year":"2014","journal-title":"Int. Orthop."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"103166","DOI":"10.1016\/j.media.2024.103166","article-title":"A survey of label\u2013noise deep learning for medical image analysis","volume":"95","author":"Shi","year":"2024","journal-title":"Med. Image Anal."},{"key":"ref_41","unstructured":"Li, M., and Zhu, C. (2024). Noisy label processing for classification: A survey. arXiv."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"2039348","DOI":"10.1080\/08839514.2022.2039348","article-title":"Effect of label noise on multi\u2013class semantic segmentation: A case study on Bangladesh marine region","volume":"36","author":"Pranto","year":"2022","journal-title":"Appl. Artif. Intell."},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"153","DOI":"10.1186\/s13643-019-1063-z","article-title":"Effects of walking speed on gait biomechanics in healthy participants: A systematic review and meta-analysis","volume":"8","author":"Fukuchi","year":"2019","journal-title":"Syst. Rev."},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"19","DOI":"10.1186\/1743-0003-2-19","article-title":"Gait variability: Methods, modeling and meaning","volume":"2","author":"Hausdorff","year":"2005","journal-title":"J. Neuroeng. Rehabil."},{"key":"ref_45","doi-asserted-by":"crossref","first-page":"80448","DOI":"10.1109\/ACCESS.2022.3192514","article-title":"Why is multiclass classification hard?","volume":"10","author":"Nowaczyk","year":"2022","journal-title":"IEEE Access"},{"key":"ref_46","doi-asserted-by":"crossref","first-page":"105344","DOI":"10.1016\/j.compbiomed.2022.105344","article-title":"Metaheuristics based COVID\u201319 detection using medical images: A review","volume":"144","author":"Riaz","year":"2022","journal-title":"Comput. Biol. Med."},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"325","DOI":"10.1016\/j.patrec.2020.11.008","article-title":"Rank consistent ordinal regression for neural networks with application to age estimation","volume":"140","author":"Cao","year":"2020","journal-title":"Pattern Recognit. Lett."},{"key":"ref_48","doi-asserted-by":"crossref","first-page":"941","DOI":"10.1007\/s10044-023-01181-9","article-title":"Deep neural networks for rank\u2013consistent ordinal regression based on conditional probabilities","volume":"26","author":"Shi","year":"2023","journal-title":"Pattern Anal. Appl."},{"key":"ref_49","doi-asserted-by":"crossref","unstructured":"Deng, J., Guo, J., Xue, N., and Zafeiriou, S. (2019). ArcFace: Additive angular margin loss for deep face recognition. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, IEEE.","DOI":"10.1109\/CVPR.2019.00482"},{"key":"ref_50","doi-asserted-by":"crossref","unstructured":"Cai, X., Luo, F., Qi, W., and Liu, H. (2022). A semi-supervised object detection algorithm based on teacher-student models with strong-weak heads. Electronics, 11.","DOI":"10.3390\/electronics11233849"},{"key":"ref_51","doi-asserted-by":"crossref","first-page":"112117","DOI":"10.1016\/j.patcog.2025.112117","article-title":"SuperCM: Improving semi-supervised learning and domain adaptation through differentiable clustering","volume":"171","author":"Singh","year":"2025","journal-title":"Pattern Recognit."},{"key":"ref_52","doi-asserted-by":"crossref","unstructured":"Fu, G., Zhang, Z., Le, W., Li, J.B., Zhu, Q.X., Niu, F.Z., Chen, H., Sun, F.Y., and Shen, Y.H. (2023). A multi\u2013scale pooling convolutional neural network for accurate steel surface defects classification. Front. Neurorobot., 17.","DOI":"10.3389\/fnbot.2023.1096083"},{"key":"ref_53","unstructured":"Zhang, X., Nie, Q., Xiao, Z., Zhao, J., Wu, X., Guo, P., Li, R., Liu, J., Wei, Y.J., and Pan, Y. (2024). Dual-view pyramid pooling in deep neural networks for improved medical image classification and confidence calibration. arXiv."},{"key":"ref_54","unstructured":"Furuta, R., and Sato, Y. (2023). Seeking flat minima with mean teacher on semi- and weakly-supervised domain generalization for object detection. arXiv."},{"key":"ref_55","doi-asserted-by":"crossref","first-page":"101886","DOI":"10.1016\/j.jksuci.2023.101886","article-title":"Multi-scale cross-layer fusion and center position network for pedestrian detection","volume":"36","author":"Liu","year":"2024","journal-title":"J. King Saud Univ.-Comput. Inf. Sci."},{"key":"ref_56","doi-asserted-by":"crossref","unstructured":"Tran, D., Wang, H., Torresani, L., Ray, J., LeCun, Y., and Paluri, M. (2018). A closer look at spatiotemporal convolutions for action recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, IEEE.","DOI":"10.1109\/CVPR.2018.00675"},{"key":"ref_57","unstructured":"Ghosh, P., Hossain, R.B., Zunaed, M., and Hasan, T. (2025). Domain generalization for improved human activity recognition in office space videos using adaptive pre-processing. arXiv."},{"key":"ref_58","doi-asserted-by":"crossref","first-page":"34","DOI":"10.1007\/s11063-025-11747-9","article-title":"A survey of data augmentation in domain generalization","volume":"57","author":"Zhong","year":"2025","journal-title":"Neural Process. Lett."}],"container-title":["Bioengineering"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2306-5354\/13\/3\/272\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T17:05:43Z","timestamp":1772125543000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2306-5354\/13\/3\/272"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,26]]},"references-count":58,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,3]]}},"alternative-id":["bioengineering13030272"],"URL":"https:\/\/doi.org\/10.3390\/bioengineering13030272","relation":{},"ISSN":["2306-5354"],"issn-type":[{"value":"2306-5354","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,26]]}}}