{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,31]],"date-time":"2026-03-31T16:34:47Z","timestamp":1774974887644,"version":"3.50.1"},"reference-count":42,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,5,6]],"date-time":"2025-05-06T00:00:00Z","timestamp":1746489600000},"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><jats:title>Background<\/jats:title><jats:p>The application of deep learning techniques in medical image analysis has shown great potential in assisting clinical diagnosis. This study focuses on the development and evaluation of deep learning models for the classification of knee joint injuries using Magnetic Resonance Imaging (MRI) data. The research aims to provide an efficient and reliable tool for clinicians to aid in the diagnosis of knee joint disorders, particularly focusing on Anterior Cruciate Ligament (ACL) tears.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>KneeXNet leverages the power of graph convolutional networks (GCNs) to capture the intricate spatial dependencies and hierarchical features in knee MRI scans. The proposed model consists of three main components: a graph construction module, graph convolutional layers, and a multi-scale feature fusion module. Additionally, a contrastive learning scheme is employed to enhance the model\u2019s discriminative power and robustness. The MRNet dataset, consisting of knee MRI scans from 1,370 patients, is used to train and validate KneeXNet.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>The performance of KneeXNet is evaluated using the Area Under the Receiver Operating Characteristic Curve (AUC) metric and compared to state-of-the-art methods, including traditional machine learning approaches and deep learning models. KneeXNet consistently outperforms the competing methods, achieving AUC scores of 0.985, 0.972, and 0.968 for the detection of knee joint abnormalities, ACL tears, and meniscal tears, respectively. The cross-dataset evaluation further validates the generalization ability of KneeXNet, maintaining its superior performance on an independent dataset.<\/jats:p><\/jats:sec><jats:sec><jats:title>Application<\/jats:title><jats:p>To facilitate the clinical application of KneeXNet, a user-friendly web interface is developed using the Django framework. This interface allows users to upload MRI scans, view diagnostic results, and interact with the system seamlessly. The integration of Grad-CAM visualizations enhances the interpretability of KneeXNet, enabling radiologists to understand and validate the model\u2019s decision-making process.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fbioe.2025.1590962","type":"journal-article","created":{"date-parts":[[2025,5,6]],"date-time":"2025-05-06T04:15:52Z","timestamp":1746504952000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["Leveraging spatial dependencies and multi-scale features for automated knee injury detection on MRI diagnosis"],"prefix":"10.3389","volume":"13","author":[{"given":"Jianhua","family":"Sun","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ye","family":"Cao","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ying","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Baoqiao","family":"Qi","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2025,5,6]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1803.01164","article-title":"The history began from alexnet: a comprehensive survey on deep learning approaches","author":"Alom","year":"2018","journal-title":"arXiv Prepr. arXiv:1803.01164"},{"key":"B2","first-page":"149","article-title":"A comparative study of existing and new deep learning methods for detecting knee injuries using the mrnet dataset","author":"Azcona","year":"2020"},{"key":"B3","first-page":"71","article-title":"Optimising knee injury detection with spatial attention and validating localisation ability","author":"Belton","year":"2021"},{"key":"B4","doi-asserted-by":"publisher","first-page":"88963","DOI":"10.1007\/s11042-024-18970-x","article-title":"Classification of cervical spine disease using convolutional neural network","volume":"83","author":"Bezabh","year":"2024","journal-title":"Multimedia Tools Appl."},{"key":"B5","doi-asserted-by":"publisher","first-page":"e1002699","DOI":"10.1371\/journal.pmed.1002699","article-title":"Deep-learning-assisted diagnosis for knee magnetic resonance imaging: development and retrospective validation of mrnet","volume":"15","author":"Bien","year":"2018","journal-title":"PLoS Med."},{"key":"B6","doi-asserted-by":"publisher","first-page":"103574","DOI":"10.1016\/j.apergo.2021.103574","article-title":"The role of machine learning in the primary prevention of work-related musculoskeletal disorders: a scoping review","volume":"98","author":"Chan","year":"2022","journal-title":"Appl. Ergon."},{"key":"B7","doi-asserted-by":"publisher","first-page":"3538","DOI":"10.1007\/s00330-020-06658-3","article-title":"Assessment of knee pain from mr imaging using a convolutional siamese network","volume":"30","author":"Chang","year":"2020","journal-title":"Eur. Radiol."},{"key":"B8","doi-asserted-by":"publisher","first-page":"980","DOI":"10.1007\/s10278-019-00193-4","article-title":"Deep learning for detection of complete anterior cruciate ligament tear","volume":"32","author":"Chang","year":"2019","journal-title":"J. digital imaging"},{"key":"B9","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/1833351.1778775","article-title":"A simple geometric model for elastic deformations","volume":"29","author":"Chao","year":"2010","journal-title":"ACM Trans. Graph. (TOG)"},{"key":"B10","doi-asserted-by":"publisher","first-page":"1414","DOI":"10.1002\/jcu.23321","article-title":"Artificial intelligence, machine learning and deep learning in musculoskeletal imaging: current applications","volume":"50","author":"D\u2019Angelo","year":"2022","journal-title":"J. Clin. Ultrasound"},{"key":"B11","doi-asserted-by":"publisher","first-page":"107570","DOI":"10.1016\/j.compbiomed.2023.107570","article-title":"Dc-aae: dual channel adversarial autoencoder with multitask learning for kl-grade classification in knee radiographs","volume":"167","author":"Farooq","year":"2023","journal-title":"Comput. Biol. Med."},{"key":"B12","first-page":"199","article-title":"Ergonomics and machine learning: wearable sensors in the prevention of work-related musculoskeletal disorders","volume-title":"Occupational and environmental safety and health IV","author":"Fernandes","year":"2022"},{"key":"B13","doi-asserted-by":"publisher","first-page":"437","DOI":"10.1002\/mrm.28111","article-title":"Automated cartilage and meniscus segmentation of knee mri with conditional generative adversarial networks","volume":"84","author":"Gaj","year":"2020","journal-title":"Magnetic Reson. Med."},{"key":"B14","first-page":"452","article-title":"Transfer learning and data augmentation in the diagnosis of knee mri","author":"Haddadian","year":"2022"},{"key":"B15","doi-asserted-by":"publisher","first-page":"378","DOI":"10.1002\/jmri.29412","article-title":"Associating knee osteoarthritis progression with temporal-regional graph convolutional network analysis on mr images","volume":"61","author":"Hu","year":"2025","journal-title":"J. Magnetic Reson. Imaging"},{"key":"B16","doi-asserted-by":"publisher","first-page":"740","DOI":"10.1002\/jmri.28284","article-title":"Automatic detection of meniscus tears using backbone convolutional neural networks on knee mri","volume":"57","author":"Hung","year":"2023","journal-title":"J. Magnetic Reson. Imaging"},{"key":"B17","doi-asserted-by":"publisher","first-page":"e220425","DOI":"10.1148\/radiol.220425","article-title":"Deep learning reconstruction enables prospectively accelerated clinical knee mri","volume":"307","author":"Johnson","year":"2023","journal-title":"Radiology"},{"key":"B18","doi-asserted-by":"publisher","first-page":"1009","DOI":"10.3390\/make3040050","article-title":"Detection and classification of knee injuries from mr images using the mrnet dataset with progressively operating deep learning methods","volume":"3","author":"Kara","year":"2021","journal-title":"Mach. Learn. Knowl. Extr."},{"key":"B19","doi-asserted-by":"publisher","first-page":"41","DOI":"10.1186\/s12891-023-06153-y","article-title":"Automatic segmentation of human knee anatomy by a convolutional neural network applying a 3d mri protocol","volume":"24","author":"Kulseng","year":"2023","journal-title":"BMC Musculoskelet. Disord."},{"key":"B20","doi-asserted-by":"publisher","first-page":"e0315809","DOI":"10.1371\/journal.pone.0315809","article-title":"Multimodal feature fusion-based graph convolutional networks for alzheimer\u2019s disease stage classification using f-18 florbetaben brain pet images and clinical indicators","volume":"19","author":"Lee","year":"2024","journal-title":"PloS one"},{"key":"B21","doi-asserted-by":"publisher","first-page":"1679","DOI":"10.3233\/jad-215497","article-title":"Predicting brain amyloid-\u03b2 pet grades with graph convolutional networks based on functional mri and multi-level functional connectivity","volume":"86","author":"Li","year":"2022","journal-title":"J. Alzheimer\u2019s Dis."},{"key":"B22","doi-asserted-by":"publisher","first-page":"276","DOI":"10.1038\/s41597-023-02196-x","article-title":"Database covering the prayer movements which were not available previously","volume":"10","author":"Mihcin","year":"2023","journal-title":"Sci. data"},{"key":"B23","doi-asserted-by":"publisher","first-page":"e190207","DOI":"10.1148\/ryai.2020190207","article-title":"Deep learning for hierarchical severity staging of anterior cruciate ligament injuries from mri","volume":"2","author":"Namiri","year":"2020","journal-title":"Radiol. Artif. Intell."},{"key":"B24","doi-asserted-by":"publisher","first-page":"2976","DOI":"10.1109\/tmi.2020.2985861","article-title":"Discriminative regularized auto-encoder for early detection of knee osteoarthritis: data from the osteoarthritis initiative","volume":"39","author":"Nasser","year":"2020","journal-title":"IEEE Trans. Med. imaging"},{"key":"B25","doi-asserted-by":"publisher","first-page":"7637","DOI":"10.1038\/s41467-024-51888-4","article-title":"Learning co-plane attention across mri sequences for diagnosing twelve types of knee abnormalities","volume":"15","author":"Qiu","year":"2024","journal-title":"Nat. Commun."},{"key":"B26","doi-asserted-by":"publisher","first-page":"1421","DOI":"10.2214\/ajr.20.23313","article-title":"Using deep learning to accelerate knee mri at 3 t: results of an interchangeability study","volume":"215","author":"Recht","year":"2020","journal-title":"Am. J. Roentgenol."},{"key":"B27","first-page":"21","article-title":"Mrnet: a software-based multicast\/reduction network for scalable tools","author":"Roth","year":"2003"},{"key":"B28","first-page":"1405","article-title":"Implementation of alexnet for classification of knee osteoarthritis","author":"Sivakumari","year":"2022"},{"key":"B29","first-page":"1","article-title":"Anterior cruciate ligament injury classification from mri scans using deep learning","author":"Tuazon","year":"2023"},{"key":"B30","doi-asserted-by":"publisher","first-page":"1053","DOI":"10.1007\/s00296-021-04916-1","article-title":"Machine learning reveals the most important psychological and social variables predicting the differential diagnosis of rheumatic and musculoskeletal diseases","volume":"42","author":"Vera Cruz","year":"2022","journal-title":"Rheumatol. Int."},{"key":"B31","doi-asserted-by":"publisher","first-page":"478","DOI":"10.3390\/life14040478","article-title":"Refined detection and classification of knee ligament injury based on resnet convolutional neural networks","volume":"14","author":"Voinea","year":"2024","journal-title":"Life"},{"key":"B32","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1007\/s40747-024-01648-0","article-title":"Multimodal heterogeneous graph fusion for automated obstructive sleep apnea-hypopnea syndrome diagnosis","volume":"11","author":"Wang","year":"","journal-title":"Complex and Intelligent Syst."},{"key":"B33","doi-asserted-by":"publisher","first-page":"102836","DOI":"10.1016\/j.inffus.2024.102836","article-title":"Autogrn: an adaptive multi-channel graph recurrent joint optimization network with copula-based dependency modeling for spatio-temporal fusion in electrical power systems","volume":"117","author":"Wang","year":"","journal-title":"Inf. Fusion"},{"key":"B34","doi-asserted-by":"publisher","first-page":"3685","DOI":"10.1109\/tcbb.2021.3126641","article-title":"sccdg: a method based on dae and gcn for scrna-seq data analysis","volume":"19","author":"Wang","year":"2021","journal-title":"IEEE\/ACM Trans. Comput. Biol. Bioinforma."},{"key":"B35","doi-asserted-by":"publisher","first-page":"1518","DOI":"10.1016\/j.acra.2023.10.036","article-title":"Automated segmentation and classification of knee synovitis based on mri using deep learning","volume":"31","author":"Wang","year":"2024","journal-title":"Acad. Radiol."},{"key":"B36","doi-asserted-by":"publisher","first-page":"512","DOI":"10.1049\/cit2.12079","article-title":"A resnet-based approach for accurate radiographic diagnosis of knee osteoarthritis","volume":"7","author":"Wang","year":"2022","journal-title":"CAAI Trans. Intell. Technol."},{"key":"B37","first-page":"1850","article-title":"View-gcn: view-based graph convolutional network for 3d shape analysis","author":"Wei","year":"2020"},{"key":"B38","doi-asserted-by":"publisher","first-page":"2620","DOI":"10.21037\/qims-21-459","article-title":"Automated knee cartilage segmentation for heterogeneous clinical mri using generative adversarial networks with transfer learning","volume":"12","author":"Yang","year":"2022","journal-title":"Quantitative Imaging Med. Surg."},{"key":"B39","doi-asserted-by":"publisher","first-page":"8841","DOI":"10.3390\/app12178841","article-title":"Gan-tl: generative adversarial networks with transfer learning for mri reconstruction","volume":"12","author":"Yaqub","year":"2022","journal-title":"Appl. Sci."},{"key":"B40","doi-asserted-by":"publisher","first-page":"112208","DOI":"10.1016\/j.knosys.2024.112208","article-title":"Federated semi-supervised representation augmentation with cross-institutional knowledge transfer for healthcare collaboration","volume":"300","author":"Yin","year":"2024","journal-title":"Knowledge-Based Syst."},{"key":"B41","doi-asserted-by":"publisher","first-page":"368","DOI":"10.1109\/tmi.2022.3206042","article-title":"Knee cartilage defect assessment by graph representation and surface convolution","volume":"42","author":"Zhuang","year":"2022","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B42","doi-asserted-by":"publisher","first-page":"104154","DOI":"10.1016\/j.bspc.2022.104154","article-title":"Multi-scale deformable transformer for multi-contrast knee mri super-resolution","volume":"79","author":"Zou","year":"2023","journal-title":"Biomed. Signal Process. Control"}],"container-title":["Frontiers in Bioengineering and Biotechnology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbioe.2025.1590962\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,6]],"date-time":"2025-05-06T04:16:06Z","timestamp":1746504966000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbioe.2025.1590962\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,5,6]]},"references-count":42,"alternative-id":["10.3389\/fbioe.2025.1590962"],"URL":"https:\/\/doi.org\/10.3389\/fbioe.2025.1590962","relation":{},"ISSN":["2296-4185"],"issn-type":[{"value":"2296-4185","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,5,6]]},"article-number":"1590962"}}