FDMRNet: A classification model for anterior cruciate ligament biomechanical injuries based on FSM and DFFM
Abstract
The lesion area in magnetic resonance imaging (MRI) of anterior cruciate ligament (ACL) injury is small, the features are difficult to focus, and the multiangle imaging features are scattered, which presents great challenges to clinicians for ACL injury. The ACL plays a critical role in maintaining knee stability. An injury can result in increased laxity, making the knee more vulnerable to further damage. This paper proposes a new neural network model, FDMRNet, which automatically focuses on the area of ACL injury and improves the accuracy of intelligent discrimination of the degree of ACL injury. Understanding the biomechanical effects of ACL injuries is crucial for developing effective rehabilitation protocols aimed at restoring normal knee function and preventing re-injury. First, FDMRNet enhances the focus of lesion features and reduces noise interference through the feature selection module (FSM), thereby improving the lesion localization ability. Secondly, the dimensional feature fusion module (DFFM) is used to fuse multi-angle features, which enhances the accuracy of the fusion representation of multi-angle features. To evaluate the performance of FDMRNet, real datasets from the Guangdong Provincial Armed Police Corps Hospital were used for model training and verification. The experimental results show that compared with the mainstream methods, the AUC (Area Under Curve), accuracy, precision, recall, and f1-score of the proposed model are improved by 2.52%, 3.17%, 5.79%, 4.14% and 4.54% respectively, which fully proves the effectiveness and accuracy of the proposed model in MRI classification of anterior cruciate ligament injury. Recognizing the biomechanical consequences of ACL injuries highlights the importance of accurate diagnosis and effective treatment strategies, which can be significantly enhanced through advanced models like FDMRNet.
References
1. Oei EHG, Nikken JJ, et al. Costs and effectiveness of a brief MRI examination of patients with acute knee injury. European Radiology. 2008; 19(2): 409-418. doi: 10.1007/s00330-008-1162-z
2. Oei EHG, Nikken JJ, Ginai AZ, et al. Acute Knee Trauma: Value of a Short Dedicated Extremity MR Imaging Examination for Prediction of Subsequent Treatment. Radiology. 2005; 234(1): 125-133. doi: 10.1148/radiol.2341031062
3. Srivastava S, Sharma G. OmniVec: Learning robust representations with cross modal sharing. arXiv; 2024.
4. Kabir H. Reduction of class activation uncertainty with background information. ArXiv; 2023.
5. Singh M, Duval Q, Alwala KV, et al. The effectiveness of MAE pre-pretraining for billion-scale pretraining. ArXiv; 2023.
6. Dosovitskiy A. An image is worth 16x16 words: Transformers for image recognition at scale. ArXiv; 2020.
7. Li M, Jiang Y, Zhang Y, et al. Medical image analysis using deep learning algorithms. Frontiers in Public Health. 2023; 11. doi: 10.3389/fpubh.2023.1273253
8. Bien N, Rajpurkar P, Ball RL, et al. Deep-learning-assisted diagnosis for knee magnetic resonance imaging: Development and retrospective validation of MRNet. Saria S, ed. PLOS Medicine. 2018; 15(11): e1002699. doi: 10.1371/journal.pmed.1002699
9. Krizhevsky A, Sutskever I, and Hinton GE. Imagenet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems. 2012; 25.
10. He K, Zhang X, Ren S, et al. Deep Residual Learning for Image Recognition. In: Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2016.
11. Štajduhar I, Mamula M, Miletić D, et al. Semi-automated detection of anterior cruciate ligament injury from MRI. Computer Methods and Programs in Biomedicine. 2017; 140: 151-164. doi: 10.1016/j.cmpb.2016.12.006
12. Zhang L, Li M, Zhou Y, et al. Deep Learning Approach for Anterior Cruciate Ligament Lesion Detection: Evaluation of Diagnostic Performance Using Arthroscopy as the Reference Standard. Journal of Magnetic Resonance Imaging. 2020; 52(6): 1745-1752. doi: 10.1002/jmri.27266
13. Germann C, Marbach G, Civardi F, et al. Deep Convolutional Neural Network–Based Diagnosis of Anterior Cruciate Ligament Tears. Investigative Radiology. 2020; 55(8): 499-506. doi: 10.1097/rli.0000000000000664
14. Namiri NK, Flament I, Astuto B, et al. Hierarchical severity staging of anterior cruciate ligament injuries using deep learning with MRI images. ArXiv; 2020.
15. Garigapati K, Blasch E, Wei J, et al. Transparent Object Tracking with Enhanced Fusion Module. In: Proceedings of the 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS); 2023.
16. Deevi SA, Lee C, Gan L, et al. RGB-X Object Detection via Scene-Specific Fusion Modules. In: Proceedings of the 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV); 2024.
17. Xie L, Xiang C, Yu Z, et al. PI-RCNN: An Efficient Multi-Sensor 3D Object Detector with Point-Based Attentive Cont-Conv Fusion Module. Proceedings of the AAAI Conference on Artificial Intelligence. 2020; 34(07): 12460-12467. doi: 10.1609/aaai.v34i07.6933
18. Chen L, Fu Y, Gu L, et al. Frequency-Aware Feature Fusion for Dense Image Prediction. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2024; 46(12): 10763-10780. doi: 10.1109/tpami.2024.3449959
19. Joze HRV, Shaban A, Iuzzolino ML, and Koishida K. MMTM: Multimodal transfer module for CNN fusion. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition; 2020.
20. Pudjihartono N, Fadason T, Kempa-Liehr AW, et al. A Review of Feature Selection Methods for Machine Learning-Based Disease Risk Prediction. Frontiers in Bioinformatics. 2022; 2. doi: 10.3389/fbinf.2022.927312
21. Halilaj E, Rajagopal A, Fiterau M, et al. Machine learning in human movement biomechanics: Best practices, common pitfalls, and new opportunities. Journal of Biomechanics. 2018; 81: 1-11. doi: 10.1016/j.jbiomech.2018.09.009
22. Chen X, He K. Exploring Simple Siamese Representation Learning. In: Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR); 2021.
23. Richemond PH, Grill J, Altché F, et al. Byol works even without batch statistics. ArXiv; 2020.
24. Kingma DP. Adam: A method for stochastic optimization. ArXiv; 2014.
25. Lin TY, Goyal P, Girshick R, et al. Focal Loss for Dense Object Detection. In: Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV); 2017.
26. Tang D, Chen J, Ren L, et al. Reviewing CAM-Based Deep Explainable Methods in Healthcare. Applied Sciences. 2024; 14(10): 4124. doi: 10.3390/app14104124
27. Javed Awan M, Mohd Rahim M, Salim N, et al. Efficient Detection of Knee Anterior Cruciate Ligament from Magnetic Resonance Imaging Using Deep Learning Approach. Diagnostics. 2021; 11(1): 105. doi: 10.3390/diagnostics11010105
28. Zhang C, Chen S, Cigdem O, et al. MR-Transformer: Vision Transformer for Total Knee Replacement Prediction Using Magnetic Resonance Imaging. ArXiv; 2024.
29. Sethi S, Reddy S, Sakarvadia M, et al. Toward non-invasive diagnosis of Bankart lesions with deep learning. ArXiv; 2024.
30. Dagli R. Astroformer: More data might not be all you need for classification. ArXiv; 2023.
31. Phong NH, Santos A, Ribeiro B. PSO-Convolutional Neural Networks With Heterogeneous Learning Rate. IEEE Access. 2022; 10: 89970-89988. doi: 10.1109/access.2022.3201142
32. Pinasthika K, Laksono BSP, Irsal RBP, et al. SparseSwin: Swin transformer with sparse transformer block. Neurocomputing. 2024; 580: 127433. doi: 10.1016/j.neucom.2024.127433
33. Qiu X, Zhu RJ, Chou Y, et al. Gated Attention Coding for Training High-Performance and Efficient Spiking Neural Networks. Proceedings of the AAAI Conference on Artificial Intelligence. 2024; 38(1): 601-610. doi: 10.1609/aaai.v38i1.27816
34. Su Z, Chen J, Pang L, et al. Adaptive Split-Fusion Transformer. In: Proceedings of the 2023 IEEE International Conference on Multimedia and Expo (ICME); 2023.
Copyright (c) 2025 Author(s)
This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright on all articles published in this journal is retained by the author(s), while the author(s) grant the publisher as the original publisher to publish the article.
Articles published in this journal are licensed under a Creative Commons Attribution 4.0 International, which means they can be shared, adapted and distributed provided that the original published version is cited.
Submit a Paper
