A Synergetic Deep Learning Framework for Enhanced Lumbar Spine MRI Segmentation Using ResNet-50 and U-Net Architectures

Afia  Zafar; Mohsina  Abid; Shahneer  Zafar; Noushin  Saba; Hina  Ayaz

doi:10.62019/sc712z60

Authors

Afia Zafar Department of Computer Science National University of Computer and Emerging Sciences, Pakistan.
Mohsina Abid Department of Computer Science National University of Computer and Emerging Sciences, Pakistan.
Shahneer Zafar Department of Computer Science Nutecch University Islamabad, Pakistan.
Noushin Saba Department of Computer Science Nutecch University Islamabad, Pakistan.
Hina Ayaz Department of Computer Science National University of Computer and Emerging Sciences, Pakistan.

DOI:

https://doi.org/10.62019/sc712z60

Abstract

Lumbar spine disorders are a leading cause of disability worldwide, significantly affecting patients’ quality of life. Traditional diagnosis methods rely on manual interpretation of MRI scans by radiologists, a process that is time-consuming, prone to human error, and susceptible to inter-observer variability. To address these limitations, this study proposes an automated segmentation approach for lumbar spine MRI images using advanced deep learning models. Specifically, U-Net and ResNet-50 architectures are employed to accurately segment critical spinal structures, thereby improving diagnostic precision and consistency. The models were trained and evaluated using a publicly available Lumbar Spine MRI dataset, and their performance was assessed using multiple metrics, including Accuracy, Precision, Recall, Intersection over Union (IoU), and Inference Time. Experimental results demonstrate that ResNet-50 outperforms U-Net in most metrics, offering higher accuracy and faster inference. This automated framework provides a reliable and efficient solution for lumbar spine analysis, with the potential to enhance clinical decision-making and reduce diagnostic delays in real-world healthcare settings.

References

Abdou, M. A. (2022). Literature review: Efficient deep neural networks techniques for medical image analysis. Neural Computing and Applications, 34(8), 5791–5812. https://doi.org/10.1007/s00521-021-XXXXX DOI: https://doi.org/10.1007/s00521-022-06960-9

Alafer, F., Siddiqi, M. H., Khan, M. S., Ahmad, I., Alhujaili, S., Alrowaili, Z., & Alshabibi, A. S. (2024). A comprehensive exploration of L-UNet approach: Revolutionizing medical image segmentation. IEEE Access, 12, 140769–140791. https://doi.org/10.1109/ACCESS.2024.XXXXX DOI: https://doi.org/10.1109/ACCESS.2024.3413038

Al-Worafi, Y. M. (2024). Musculoskeletal diseases: Causes and risk factors in developing countries. In Handbook of medical and health sciences in developing countries: Education, practice, and research (pp. 1–21). Springer International Publishing. https://doi.org/10.1007/978-3-031-XXXXX-X_1 DOI: https://doi.org/10.1007/978-3-030-74786-2_324-1

Azad, R., Aghdam, E. K., Rauland, A., Jia, Y., Avval, A. H., Bozorgpour, A., ... & Merhof, D. (2024). Medical image segmentation review: The success of U-Net. IEEE Transactions on Pattern Analysis and Machine Intelligence. https://doi.org/10.1109/TPAMI.2024.XXXXX DOI: https://doi.org/10.1109/TPAMI.2024.3435571

Baloch, L., Sajid, A., Dewi, C., Christanto, H. J., & Zafar, A. (2024). Deep learning-based medical image reconstruction: Overview, analysis, and challenges. Revue d'Intelligence Artificielle, 38(2), 417. https://doi.org/10.18280/ria.3802XXX DOI: https://doi.org/10.18280/ria.380205

Barbhuiya, R. K., & Paul, C. (2025). Deep learning transformations in medical imaging: Advancements in brain tumor, retinal vessel, and inner ear segmentation. In Deep learning applications in medical image segmentation: Overview, approaches, and challenges (pp. 113–132). https://doi.org/10.1007/978-3-031-XXXXX-X_6 DOI: https://doi.org/10.1002/9781394245369.ch5

Chang, D., Lui, A., Matsoyan, A., Safaee, M., Aryan, H., & Ames, C. (2024). Comparative review of the socioeconomic burden of lower back pain in the United States and globally. Neurospine, 21(2), 487. https://doi.org/10.14245/ns.XXXXX DOI: https://doi.org/10.14245/ns.2448372.186

Gupta, R. S., Lal, B., Bhagat, A. C., & Alagarsamy, R. (2024). Medical imaging for patient-specific implants. In Biomedical implants (pp. 39–60). CRC Press. DOI: https://doi.org/10.1201/9781003375098-4

Katz, J. N., Zimmerman, Z. E., Mass, H., & Makhni, M. C. (2022). Diagnosis and management of lumbar spinal stenosis: A review. JAMA, 327(17), 1688–1699. https://doi.org/10.1001/jama.2022.XXXXX DOI: https://doi.org/10.1001/jama.2022.5921

Kumar, A. A., & Kesavadas, C. (2025). Potential of MRI in clinical medicine. In Multimodal biomedical imaging techniques (pp. 271–301). Springer Nature Singapore. https://doi.org/10.1007/978-981-XXXXX-X_10 DOI: https://doi.org/10.1007/978-981-96-1124-9_12

Muhaimil, A., Pendem, S., Sampathilla, N., P. S., P., Nayak, K., Chadaga, K., ... & Shirlal, A. (2024). Role of artificial intelligence model in prediction of low back pain using T2-weighted MRI of lumbar spine. F1000Research, 13, 1035. https://doi.org/10.12688/f1000research.152879.1 DOI: https://doi.org/10.12688/f1000research.154680.2

Panahi, O. (2025). Deep learning in diagnostics. Journal of Medical Discoveries, 2(1), 1–6. https://doi.org/10.1234/jmd.2025.XXXXX

Santiago, F. R., Ramos-Bossini, A. J. L., Wáng, Y. X. J., Barbero, J. P. M., Espinosa, J. G., & Martínez, A. M. (2022). The value of magnetic resonance

imaging and computed tomography in the study of spinal disorders. Quantitative Imaging in Medicine and Surgery, 12(7), 3947. https://doi.org/10.21037/qims-XXXX

Talaat, F. M., El-Sappagh, S., Alnowaiser, K., & Hassan, E. (2024). Improved prostate cancer diagnosis using a modified ResNet50-based deep learning architecture. BMC Medical Informatics and Decision Making, 24(1), 23. https://doi.org/10.1186/s12911-024-XXXXX DOI: https://doi.org/10.1186/s12911-024-02419-0

Zafar, A., Saba, N., Arshad, A., Alabrah, A., Riaz, S., Suleman, M., ... & Nadeem, M. (2024). Convolutional neural networks: A comprehensive evaluation and benchmarking of pooling layer variants. Symmetry, 16(11), 1516. https://doi.org/10.3390/sym16111516 DOI: https://doi.org/10.3390/sym16111516

Zhao, T., Sun, Z., Guo, Y., Sun, Y., Zhang, Y., & Wang, X. (2023). Automatic renal mass segmentation and classification on CT images based on 3D U-Net and ResNet algorithms. Frontiers in Oncology, 13, 1169922. https://doi.org/10.3389/fonc.2023.1169922 DOI: https://doi.org/10.3389/fonc.2023.1169922