Article Sidebar

PDF
Issue: No. 98 (2026): CHÀO MỪNG NGÀY KHOA HỌC, CÔNG NGHỆ VÀ ĐỔI MỚI SÁNG TẠO VIỆT NAM NĂM 2026
Section: Article
DOI: 10.58490/ctjump.2026i98.4793
Date Published: 25/05/2026
Views 0
Downloads 0
How to Cite
Van, M. K., Truong, N. K. H., Tran, Q. S., Nguyen, T. T., Nguyen, M. L., Nguyen, P. T., & Nguyen, T. T. (2026). APPLICATION OF CONVOLUTIONAL NEURAL NETWORKS IN DIAGNOSING ANTERIOR CRUCIATE LIGAMENT INJURIES OF THE KNEE JOINT BASED ON MAGNETIC RESONANCE IMAGING. Cantho Journal of Medicine and Pharmacy, 98, 1-7. https://doi.org/10.58490/ctjump.2026i98.4793

APPLICATION OF CONVOLUTIONAL NEURAL NETWORKS IN DIAGNOSING ANTERIOR CRUCIATE LIGAMENT INJURIES OF THE KNEE JOINT BASED ON MAGNETIC RESONANCE IMAGING

Minh Khen Van1, Nguyen Khanh Hung Truong2, Quang Son Tran1, Tam Tu Nguyen3, Minh Luan Nguyen3, Phu Toan Nguyen3, Thanh Tan Nguyen1,
1 Can Tho University of Medicine and Pharmacy
2 Cho Ray Hospital
3 Can Tho Central General Hospital

Main Article Content

Abstract

Background: Interpreting magnetic resonance imaging  scans for diagnosing anterior cruciate ligament injuries is currently time-consuming and highly dependent on the physician's experience. Convolutional Neural Networks, with their capability for automatic image feature extraction, hold promise in optimizing this workflow. Objective: To evaluate the performance of convolutional neural networks in diagnosing knee anterior cruciate ligament injuries based on magnetic resonance imaging scans. Materials and methods: A cross-sectional descriptive study was conducted on 400 patients with knee anterior cruciate ligament injuries, evaluated on non-contrast 1.5 Tesla magnetic resonance imaging (MRI) scans in the sagittal plane. The data distribution included: 341 training images, 31 internal validation images, and 28 real-world validation images from Can Tho Central General Hospital and Can Tho University of Medicine and Pharmacy Hospital. Three convolutional neural network models (EfficientNet, ResNet-50, DenseNet121) were evaluated and benchmarked against the gold standard of knee arthroscopic surgery results. Results: The convolutional neural network models achieved outstanding performance on the real-world validation set, with an accuracy of >85% and an Area Under the Curve (AUC) of >0.95. GradCAM heatmaps accurately and logically localized the damaged anatomical regions. The inference speed ranged from 4.4 to 15.8 images/second, well satisfying the requirements for rapid medical data analysis. Conclusion: Convolutional neural networks serve as an accurate diagnostic solution for anterior cruciate ligament tears, demonstrating strong generalization capabilities and scientific lesion localization. They hold great promise in becoming a powerful assistive tool for radiologists. 

Keywords

CNN, artificial intelligence, anterior cruciate ligament tear, knee MRI

Article Details

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

References

1. Namiri N.K., Flament I., Astuto B., Shah R., Tibrewala R., et al. Hierarchical severity staging of anterior cruciate ligament injuries using deep learning with MRI images. Radiol Artif Intell. 2020. 2(4), e190207, doi: 10.1148/ryai.2020190207.
2. Bien N., Rajpurkar P., Ball R.L., Irvin J., Park A., et al. Deep-learning-assisted diagnosis for knee magnetic resonance imaging: development and retrospective validation of MRNet. PLoS Med. 2018. 15(11), e1002699, doi: 10.1371/journal.pmed.1002699.
3. Liu F., Guan B., Zhou Z., Samsonov A., Rosas H., et al. Fully automated diagnosis of anterior cruciate ligament tears on knee MR images by using deep learning. Radiol Artif Intell. 2019. 1(3), e180091, doi: 10.1148/ryai.2019180091.
4. Kelly C.J., Karthikesalingam A., Suleyman M., Corrado G., and King D. Key challenges for delivering clinical impact with artificial intelligence. BMC Med, 2019. 17(1), 195, doi: 10.1186/s12916-019-1426-2.
5. Štajduhar I., Mamula M., Miletić D., and Ünal G. Semi-automated detection of anterior cruciate ligament injury from MRI. Comput Methods Programs Biomed. 2017. 140, 151-164, doi: 10.1016/j.cmpb.2016.12.006.
6. Selvaraju R.R., Cogswell M., Das A., Vedantam R., Parikh D., and Batra D. CAM: visual explanations from deep networks via gradient-based localization. Proceedings of the IEEE International Conference on Computer Vision. 2017. 618-626, doi: 10.1109/ICCV.2017.74.
7. Zech J.R., Badgeley M.A., Liu M., Costa A.B., Titano J.J., and Oermann E.K. Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: a cross-sectional study. PLoS Med. 2018. 15(11), e1002683, doi: 10.1371/journal.pmed.1002683.
8. Huang G., Liu Z., Van Der Maaten L., and Weinberger K.Q. Densely connected convolutional networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017. 4700-4708, doi: 10.1109/CVPR.2017.243.
9. He K., Zhang X., Ren S., and Sun J. Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016. 770-778, doi: 10.1109/CVPR.2016.90.
10. Ghassemi M., Oakden-Rayner L., and Beam A.L. The false hope of current approaches to explainable artificial intelligence in health care. Lancet Digit Health. 2021. 3(11), e745-e750, doi: 10.1016/S2589-7500(21)00208-9.
11. van Leeuwen K.G., Schalekamp S., Rutten M.J.C.M., van Ginneken B., and de Rooij M. Artificial intelligence in radiology: 100 commercially available products and their scientific evidence. Eur Radiol. 2021. 31(6), 3797-3804, doi: 10.1007/s00330-021-07892-z.