Accurate Segmentation Method of Sacroiliac Joint CT Images Based on Improved U-Net

Ying Jing; Yingchen  Xu

doi:10.71204/0ek2wm31

Authors

Ying Jing Taiyuan Normal University Author
Yingchen Xu Taiyuan Normal University Author

DOI:

https://doi.org/10.71204/0ek2wm31

Keywords:

Sacroiliac Joint Segmentation, CT Image, 3D U-Net, SE Module, Channel Attention, Medical Image Analysis

Abstract

To address the challenges of automatic segmentation in sacroiliac joint CT images caused by complex bone structures and narrow joint spaces, this paper proposes an improved 3D U-Net architecture that significantly enhances segmentation accuracy through embedding a Squeeze-and-Excitation (SE) channel attention module in the bottleneck layer. As sacroiliac joint segmentation is critical for early diagnosis of ankylosing spondylitis (AS), yet existing automated methods struggle with low contrast joint spaces and heterogeneous bone densities in CT images, our method utilizes the SE module's dynamic channel recalibration mechanism to enhance key feature channel weights in the deepest semantic bottleneck layer. This approach effectively resolves traditional methods' issues of over-segmentation in high-density bone cortex areas and under-segmentation in joint spaces. Evaluation on a clinical dataset of 40 sacroiliac joint CT scans demonstrates superior performance, achieving a 91.4% Dice coefficient and 84.3% IoU, representing improvements of 1.1% and 1.7% over the baseline 3D U-Net, respectively, with particular advantages in segmenting joint surface erosion areas in AS patients.

References

Chen, X., Yao, L., & Zhang, Y. (2020). Residual attention U-Net for automated multi-class segmentation of COVID-19 chest CT images. arXiv. http://arxiv.org/abs/2004.05645

Gao, J., & Wang, X. (2025). Medical image adaptive segmentation method based on SAM. Computer Applications and Software, 1–8. Retrieved May 6, 2025.

Gartenberg, A., Nessim, A., & Cho, W. (2021). Sacroiliac joint dysfunction: Pathophysiology, diagnosis, and treatment. European Spine Journal, 30, 2936–2943.

Iek, zgün, Abdulkadir, A., Lienkamp, S. S., Brox, T., & Ronneberger, O. (2016). 3d u-net: learning dense volumetric segmentation from sparse annotation. Springer, Cham.

Isensee, F., Petersen, J., Kohl, S. A. A., et al. (2021). nnU-Net: A self-configuring method for deep learning-based biomedical image segmentation. Nature Methods, 18(2), 203–211.

Jin, T., Wang, Z., & Li, Z. (2025). Medical liver computed tomography image segmentation algorithm based on multi-scale attention U-Net. Journal of Harbin Engineering University, 46(3), 529–539.

Kojima, I. , Nogami, S. , Hitachi, S. , Shimada, Y. , Ezoe, Y. , & Yokoyama-Sato, Y. , et al. (2024). Temporomandibular joint ankylosis suspected to be associated with ankylosing spondylitis based on cervical computed tomography images: a pictorial essay. Imaging Science in Dentistry, 54(2), 201-206.

Liu, X., Yu, H., Li, B., et al. (2021). Vertebra CT image segmentation method based on improved U-Net model. Journal of Harbin Institute of Technology, 26(3), 58–64.

Milletari, F., Navab, N., & Ahmadi, S.-A. (2016). V-Net: Fully convolutional neural networks for volumetric medical image segmentation. arXiv. http://arxiv.org/abs/1606.04797

Oktay, O., Schlemper, J., Folgoc, L. L., et al. (2018). Attention U-Net: Learning where to look for the pancreas. Retrieved May 29, 2023.

Ren, X., Zhao, M., Hu, W., et al. (2025). Lightweight medical image segmentation method based on local context-guided feature deep fusion. Journal of Zhengzhou University (Science Edition), 1–8. Retrieved May 6, 2025,

Ronneberger, O., Fischer, P., & Brox, T. . (2015). U-Net: Convolutional Networks for Biomedical Image Segmentation. International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer International Publishing.

Wang, R., Lei, T., Cui, R., et al. (2022). Medical image segmentation using deep learning: A survey. IET Image Processing, 16(5), 1243–1267.

Xiong, F., & Wei, Y. (2024). Optimization of segmentation model based on maximization information fusion and its application in nuclear image analysis. Multimedia Systems, 30(2), 1-17.

Yuan, F., Zuo, Z., Jiang, Y., Shu, W., Tian, Z., Ye, C., Yang, J., Mao, Z., Huang, X., Gu, S., & Peng, Y. (2025). AI-Driven Optimization of Blockchain Scalability, Security, and Privacy Protection. Algorithms, 18(5), 263.

Zhang, K., Li, X., Wang, Y., et al. (2023). Automatic image segmentation and grading diagnosis of sacroiliitis associated with AS using a deep convolutional neural network on CT images. Journal of Digital Imaging, 36(5), 2025–2034.

Zou, Q., & Liu, F. (2024). Improved U-Net fusion dilated convolution algorithm for liver computed tomography image segmentation. Laboratory Research and Exploration, 43(9), 19–24.