Semi-Supervised Image Segmentation via Heterogeneous Dual-Teacher Mechanism and Multi-Directional Feature Modeling
Source abstract
Abstract
To address label scarcity and anisotropic structural distortion from unidirectional modeling in semi-supervised cardiac MRI segmentation, we propose DTCL-Swin-UMamba. It features a novel Swin-UMamba backbone and a Multi-directional Vision-based Spatial Synthesis Module (M-VSS). To mitigate the noise sensitivity of traditional Mamba, the weight-shared M-VSS executes eight-directional parallel spatial scanning to build isotropic feature representations, enhancing thin-wall and direction-dependent structure segmentation. The Swin-UMamba backbone fuses global context from Swin Transformer with linear-complexity long-range dependency learning from Mamba. Furthermore, a heterogeneous dual-teacher training strategy—comprising an EMA-updated Swin Transformer teacher and a frozen MS-UNet teacher—provides complementary global and local structural supervision for the student network. Evaluated on the public ACDC and LA datasets, our framework achieves state-of-the-art performance. On ACDC with only 10% labeled images, it achieves Dice scores of 92.15%, 89.57%, and 87.25% for the left ventricle, right ventricle, and myocardium, respectively, with consistent gains demonstrated on the LA dataset.
Source: crossref · Quality: baseline
Affiliations
School of Computer Science and Communication Engineering, Jiangsu University, Zhenjiang 212013, China
References (20)
Show source-provided references
1. Ronneberger, O., Fischer, P., and Brox, T. (2015). U-Net: Convolutional networks for biomedical image segmentation. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015, Munich, Germany, 5–9 October 2015, Springer.10.1007/978-3-319-24574-4_28 ↗
2. nnU-Net: A self-configuring method for deep learning-based biomedical image segmentation10.1038/s41592-020-01008-z ↗
3. Medical image segmentation using deep learning: A survey10.1049/ipr2.12419 ↗
4. DPF-UNet: A CNN-Swin Transformer fusion network for 3D brain tumor segmentation in MRI images10.1007/s11227-025-07339-y ↗
5. Cao, H., Wang, Y., Chen, J., Jiang, D., Zhang, X., Tian, Q., and Wang, M. (2022). Swin-Unet: Unet-like pure transformer for medical image segmentation. Proceedings of the European Conference Computer Vision (ECCV), Tel Aviv, Israel, 23–27 October 2022, Springer.10.1007/978-3-031-25066-8_9 ↗
6. Chen, J., Lu, Y., Yu, Q., Luo, X., Adeli, E., Wang, Y., Lu, L., Yuille, A.L., and Zhou, Y. (2021). TransUNet: Transformers make strong encoders for medical image segmentation. arXiv.
7. Liu, Z., Hu, H., Lin, Y., Yao, Z., Xie, Z., Wei, Y., Ning, J., Cao, Y., Zhang, Z., and Dong, L. (2022, January 18–24). Swin Transformer V2: Scaling up capacity and resolution. Proceedings of the IEEE/CVF Conference Computer Vision Pattern Recognition (CVPR), New Orleans, LA, USA.10.1109/CVPR52688.2022.01170 ↗
8. Wang, Z., Zheng, J.-Q., Zhang, Y., Cui, G., and Li, L. (2024). Mamba-UNet: UNet-like pure visual Mamba for medical image segmentation. arXiv.10.2139/ssrn.5097998 ↗
9. Uncertainty-aware pseudo-label consistency for semi-supervised medical image segmentation10.1016/j.bspc.2022.104203 ↗
10. Zhang, Y., Yang, J., Liu, Y., Cheng, Y., and Qi, Y. (2024, January 3–6). SemiSAM: Enhancing semi-supervised medical image segmentation via SAM-assisted consistency regularization. Proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Lisbon, Portugal.10.1109/BIBM62325.2024.10821951 ↗
11. Zhang, P., Guo, A.J.X., Luo, S., Han, Z., and Guo, L. (2025). Dual teacher-student learning for semi-supervised medical image segmentation. arXiv.
12. Liu, J., Desrosiers, C., and Zhou, Y. (2022). Semi-supervised medical image segmentation using cross-model pseudo-supervision with shape awareness and local context constraints. Proceedings of the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), Singapore, 18–22 September 2022, Springer.10.1007/978-3-031-16452-1_14 ↗
13. Gu, A., and Dao, T. (2024). Mamba: Linear-time sequence modeling with selective state spaces. arXiv.
14. Zhu, L., Liao, B., Zhang, Q., Wang, X., Liu, W., and Wang, X. (2024). Vision Mamba: Efficient visual representation learning with bidirectional state space model. arXiv.
15. Ma, C., and Wang, Z. (2024). Semi-Mamba-UNet: Pixel-Level Contrastive and Pixel-Level Cross-Supervised Visual Mamba-based UNet for Semi-Supervised Medical Image Segmentation. arXiv.10.1016/j.knosys.2024.112203 ↗
16. Wang, Z., and Ma, C. (2024). Weak-Mamba-UNet: Visual Mamba makes CNN and ViT work better for scribble-based medical image segmentation. arXiv.
17. Zhang, M., Yu, Y., Jin, S., Gu, L., Ling, T., and Tao, X. (2024). VM-UNet V2: Rethinking vision Mamba UNet for medical image segmentation. Lecture Notes in Computer Science (LNCS), Springer.10.1007/978-981-97-5128-0_27 ↗
18. Liu, Y., Tian, Y., Zhao, Y., Yu, H., Xie, L., Wang, Y., Ye, Q., Jiao, J., and Liu, Y. (2024). VMamba: Visual state space model. arXiv.10.52202/079017-3273 ↗
19. MCMT: A multi-teacher cross-model training framework for semi-supervised cardiac image segmentation
20. PLMT: Pseudo-labeling with multi-teacher collaboration for semi-supervised medical image segmentation