Post-Training Network Compression for 3D Medical Image Segmentation: Reducing Computational Efforts via Tucker Decomposition
Abstract
Tucker decomposition–based post-training compression of the TotalSegmentator model reduced the computational demand of three-dimensional convolution models, maintained high segmentation accuracy, and enhanced model inference speed on less powerful hardware.
Purpose
To investigate whether the computational effort of three-dimensional CT-based multiorgan segmentation with TotalSegmentator can be reduced via Tucker decomposition–based network compression.
Materials and Methods
In this retrospective study, Tucker decomposition was applied to the convolutional kernels of the TotalSegmentator model, an nnU-Net model trained on a comprehensive CT dataset for automatic segmentation of 117 anatomic structures. The proposed approach reduced the floating-point operations and memory required during inference, offering an adjustable trade-off between computational efficiency and segmentation quality. This study used the publicly available TotalSegmentator dataset containing 1228 segmented CT scans and a test subset of 89 CT scans and used various downsampling factors to explore the relationship between model size, inference speed, and segmentation accuracy. Segmentation performance was evaluated using the Dice score.
Results
The application of Tucker decomposition to the TotalSegmentator model substantially reduced the model parameters and floating-point operations across various compression ratios, with limited loss in segmentation accuracy. Up to 88.17% of the model’s parameters were removed, with no evidence of differences in performance compared with the original model for 113 of 117 classes after fine-tuning. Practical benefits varied across different graphics processing unit architectures, with more distinct speedups on less powerful hardware.
Conclusion
The study demonstrated that post hoc network compression via Tucker decomposition presents a viable strategy for reducing the computational demand of medical image segmentation models without substantially impacting model accuracy.
Keywords: Deep Learning, Segmentation, Network Compression, Convolution, Tucker Decomposition
Supplemental material is available for this article.
© RSNA, 2025
References
- 1. . Medical image segmentation: a review. Int J Comput Sci Mobile Comput 2013;2(1):22–27. https://ijcsmc.com/docs/papers/january2013/V2I1201306.pdf.
- 2. . Medical image segmentation using deep learning: a survey. IET Image Process 2022;16(5):1243–1267.
- 3. . The multimodal brain tumor image segmentation benchmark (BRATS). IEEE Trans Med Imaging 2015;34(10):1993–2024.
- 4. . The medical segmentation decathlon. Nat Commun 2022;13(1):4128.
- 5. . U-Net: Convolutional Networks for Biomedical Image Segmentation. In: Navab N, Hornegger J, Wells W, Frangi A, eds.
Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. MICCAI 2015. Lecture Notes in Computer Science , vol 9351. Springer, 2015; 234–241. - 6. . V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation. In:
2016 Fourth International Conference on 3D Vision (3DV) . IEEE, 2016; 565–571. - 7. . Medical image segmentation review: the success of U-Net. IEEE Trans Pattern Anal Mach Intell 2024;46(12):10076–10095.
- 8. . TotalSegmentator: Robust Segmentation of 104 Anatomic Structures in CT Images. Radiol Artif Intell 2023;5(5):e230024.
- 9. . nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat Methods 2021;18(2):203–211.
- 10. . TotalSegmentator: A Gift to the Biomedical Imaging Community. Radiol Artif Intell 2023;5(5):e230235.
- 11. . Some mathematical notes on three-mode factor analysis. Psychometrika 1966;31(3):279–311.
- 12. . Compression of Deep Convolutional Neural Networks for Fast and Low Power Mobile Applications. ArXiv 1511.06530 [preprint] https://arxiv.org/abs/1511.06530. Posted November 20, 2015. Accessed January 2024.
- 13. . Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J R Stat Soc Series B Stat Methodol 2011;73(1):3–36.
- 14. . Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decomposition. ArXiv 1412.6553 [preprint] https://arxiv.org/abs/1412.6553. Posted December 19, 2014. Accessed January 2024.
- 15. . Tensor rank and the ill-posedness of the best low-rank approximation problem. SIAM J Matrix Anal Appl 2008;30(3):1084–1127.
- 16. . Model Compression with NAS and Knowledge Distillation for Medical Image Segmentation. In:
DSIT 2021: 2021 4th International Conference on Data Science and Information Technology . ACM,2021; 173–176. - 17. . Efficient medical image segmentation based on knowledge distillation. IEEE Trans Med Imaging 2021;40(12):3820–3831.
- 18. . Sculpting Efficiency: Pruning Medical Imaging Models for On-Device Inference. ArXiv 2309.05090 [preprint] https://arxiv.org/abs/2309.05090. Posted September 10, 2023. Accessed January 2024.
- 19. . Global analytic solution of fully-observed variational Bayesian matrix factorization. J Mach Learn Res 2013;14(1):1–37.
Article History
Received: June 12 2024Revision requested: Aug 3 2024
Revision received: Dec 13 2024
Accepted: Dec 17 2024
Published online: Jan 15 2025
