Nowadays, significant progress has been made in Mamba-based medical image segmentation models, however, there is an imbalance between global and local information extraction in these models, we propose a multi-scale synergistic mechanism aiming at the feature extraction stage to be able to extract global and local information effectively. We use a hierarchical receptive field attention strategy in the feature extraction stage to extract global and local information through different receptive fields. In addition, we introduced spatial and channel attention to emphasize on the image recovery stage to ensure that global and local information is not lost. We evaluated the method on ISIC2017, ISIC2018 and Synapse.The Dice similarity coefficient and intersection over union (IoU) metrics on ISIC17, ISIC18 are 89.76 and 81.41, 90.25 and 82.24, respectively. Our model achieves state-of-the-art(sota) results on the ISIC dataset compared to the Mamba approach. On Synapse the Dice and Average Hausdorff distance (HD95) are 80.17 and 14.21. The experimental results show that the proposed method has good generalization ability and robustness, which provides important support for clinical diagnosis and treatment.

Mamba;Medical image segmentation;Multi-scale

[1] Ronneberger O, Fischer P, Brox T: U-net: Convolutional networks for biomedical image segmentation. In: Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015, Munich, Germany: Springer International Publishing, 234-241, 2015. https://doi.org/10.1007/978-3-319-24574-4_28. [2] Vaswani A, Shazeer N, Parmar N, et al: Attention is all you need. Adv Neural Inf Process Syst 30:5998-6008, 2017. https://doi.org/10.48550/arXiv.1706.03762. [3] Chen J, Lu Y, Yu Q, et al: TransUNet: Rethinking the U-Net architecture design for medical image segmentation through the lens of transformers. Medical Image Analysis, 97, 103280, 2024. https://doi.org/10.1016/j.media.2024.103280. [4] Zhang Y, Liu H, Hu Q: TransFuse: Fusing Transformers and CNNs for medical image segmentation. In: Medical Image Computing and Computer Assisted Intervention–MICCAI 2021, Strasbourg, France: Springer International Publishing, 14-24, 2021. https://doi.org/10.1007/978-3-030-87193-2_2 . [5] Liu Y, Xiao T, Gong M, et al: Vmamba: Visual state space model. arXiv preprint arXiv:2401.10166, January 18, 2024. https://doi.org/10.48550/arXiv.2401.10166. [6] Ruan J, Xiang S: VM-UNet: Vision mamba UNet for medical image segmentation. arXiv preprint arXiv:2402.02491, February 5, 2024. https://doi.org/10.48550/arXiv.2402.02491. [7] Si Y, Xu H, Zhu X, et al. SCSA: Exploring the synergistic effects between spatial and channel attention. arXiv preprint arXiv:2407.05128, 2024. https://doi.org/10.48550/arXiv.2407.05128. [8] Zhou Z, Rahman Siddiquee MM, Tajbakhsh N, et al: UNet++: A nested u-net architecture for medical image segmentation. In: Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support, Granada, Spain: Springer International Publishing, 3-11, 2018. https://doi.org/10.1007/978-3-030-00889-5_1. [9] Cao, H. et al. Swin-Unet: Unet-Like Pure Transformer for Medical Image Segmentation. In: Karlinsky, L., Michaeli, T., Nishino, K. (eds) Computer Vision – ECCV 2022 Workshops. ECCV 2022. Lecture Notes in Computer Science, vol 13803. Springer, Cham. https://doi.org/10.1007/978-3-031-25066-8_9. [10] Gu, Albert, Karan Goel, and Christopher Ré. "Efficiently modeling long sequences with structured state spaces." arXiv preprint arXiv:2111.00396 (2021). https://doi.org/10.48550/arXiv.2111.00396. [11] Ma, Jun, Feifei Li, and Bo Wang. "U-mamba: Enhancing long-range dependency for biomedical image segmentation." arXiv preprint arXiv:2401.04722 (2024). https://doi.org/10.48550/arXiv.2401.04722. [12] Xing, Zhaohu, et al. "Segmamba: Long-range sequential modeling mamba for 3d medical image segmentation." International Conference on Medical Image Computing and Computer-Assisted Intervention. Cham: Springer Nature Switzerland, 2024. https://doi.org/10.1007/978-3-031-72111-3_54. [13] Ruan J, Xiang S: VM-UNet: Vision mamba UNet for medical image segmentation. arXiv preprint arXiv:2402.02491, February 5, 2024. https://doi.org/10.48550/arXiv.2402.02491 . [14] Liu Y, Xiao T, Gong M, et al: Vmamba: Visual state space model. arXiv preprint arXiv:2401.10166, January 18, 2024. https://doi.org/10.48550/arXiv.2401.10166. [15] Si, Yunzhong, et al. "SCSA: Exploring the synergistic effects between spatial and channel attention." arXiv preprint arXiv:2407.05128 (2024). https://doi.org/10.48550/arXiv.2407.05128. [16] Gao, Yunhe, et al. "A data-scalable transformer for medical image segmentation: architecture, model efficiency, and benchmark." arXiv preprint arXiv:2203.00131 (2022). https://doi.org/10.48550/arXiv.2203.00131. [17] Ruan J, Xiao C, Tong H, et al: MALUNet: A multi-attention and light-weight UNet for skin lesion segmentation. In: 2022 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Las Vegas, NV: IEEE, 264-269, 2022. https://doi.org/10.1109/BIBM55620.2022.9995040. [18] Wei J, Hu Y, Zhang R, et al: Shallow attention network for polyp segmentation. In: Medical Image Computing and Computer Assisted Intervention–MICCAI 2021, Strasbourg, France: Springer International Publishing, 699-708, 2021. https://doi.org/10.1007/978-3-030-87193-2_66 . [19] Milletari, Fausto, Nassir Navab, and Seyed-Ahmad Ahmadi. "V-net: Fully convolutional neural networks for volumetric medical image segmentation." 2016 fourth international conference on 3D vision (3DV). Ieee, 2016. https://doi.org/10.1109/3DV.2016.79. [20] Alom, Md Zahangir, et al. "Recurrent residual U-Net for medical image segmentation." Journal of medical imaging 6.1 (2019): 014006-014006. [21] Azad, Reza, et al. "Transnorm: Transformer provides a strong spatial normalization mechanism for a deep segmentation model." IEEE access 10 (2022): 108205-108215. https://doi.org/10.1109/ACCESS.2022.3211501. [22] Liu Z, Lin Y, Cao Y, et al: Swin Transformer: Hierarchical vision Transformer using shifted windows. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, Montreal, Canada: IEEE, 10012-10022, 2021. https://doi.org/10.48550/arXiv.2103.14030 . [23] Azad, Reza, et al. "Transdeeplab: Convolution-free transformer-based deeplab v3+ for medical image segmentation." International Workshop on PRedictive Intelligence In MEdicine. Cham: Springer Nature Switzerland, 2022. https://doi.org/10.1007/978-3-031-16919-9_9. [24] Wang, Haonan, et al. "Uctransnet: rethinking the skip connections in u-net from a channel-wise perspective with transformer." Proceedings of the AAAI conference on artificial intelligence. Vol. 36. No. 3. 2022. https://doi.org/10.1609/aaai.v36i3.20144. [25] Wang, Hongyi, et al. "Mixed transformer u-net for medical image segmentation." ICASSP 2022-2022 IEEE international conference on acoustics, speech and signal processing (ICASSP). IEEE, 2022. https://doi.org/10.1109/ICASSP43922.2022.9746172.

DOI： https://doi.org/10.62639/sspjiss28.20250203

Published： 2025