基于多尺度注意力的器官图像分割方法

doi:10.16088/j.issn.1001-6600.2023112501

摘要/Abstract

摘要： 图像分割技术是医学图像研究领域的重要分支,该技术有助于医生对癌症的诊断和治疗。为进一步提高图像分割的精确度,本文提出一种多尺度轴向注意力模型MAU-Net(multi-scale axial attention U-Net)用于器官图像分割。首先,该模型在编码器阶段采用深度残差网络来提取图像特征,提高模型泛化能力;其次,使用像素块融合模块(pixels fuse module, PFM)对编码器的特征信息进行再编码和线性增强,增强特征的位置信息提取能力;最后,在解码器间加入多分支轴向注意力模块(multi-branch axial attention module, MAM)来捕捉上下文信息,从而增强模型识别关键特征信息能力。在Synapse、ACDC、SegTHOR等多个多器官图像数据集上的实验结果表明, MAU-Net在器官识别和边缘预测方面均能实现更好的效果。

关键词: 图像分割, 器官分割, 注意力机制, 胸腹部器官, 深度学习

Abstract: Image segmentation technology is an important branch in the field of medical image research, and this technology helps doctors diagnose and treat cancer. In order to further improve the accuracy of image segmentation, a multi-scale axial attention model MAU-Net (multi-scale axial attention U-Net) is proposed in this paper for organ segmentation. Firstly, the model uses a deep residual network to extract image features in the encoder stage to improve the model’s generalization ability. Secondly, a pixel fusion module (PFM) is added to the decoder to enhance the ability to extract feature position information by re-encoding and linearly enhancing the feature information of the encoder. Finally, a multi-branch axial attention module (MAM) is added between the decoders to capture contextual information and enhance the ability to identify key feature information. Experimental results on multiple multi-organ image data sets such as Synapse, ACDC, and SegTHOR show that MAU-Net can achieve better results in both organ recognition and edge prediction.

Key words: image segmentation, organ segmentation, attention mechanism, thoracic and abdominal organs, deep learning

中图分类号: TP391.41

卢家辉, 陈庆锋, 王文广, 余谦, 何乃旭, 韩宗钊. 基于多尺度注意力的器官图像分割方法[J]. 广西师范大学学报（自然科学版）, 2024, 42(6): 138-148.

LU Jiahui, CHEN Qingfeng, WANG Wenguang, YU Qian, HE Naixu, HAN Zongzhao. Multi-scale Attention Learning for Abdomen Multi-organ Image Segmentation[J]. Journal of Guangxi Normal University(Natural Science Edition), 2024, 42(6): 138-148.

参考文献

[1] 郭雯, 鞠忠建, 吴青南, 等. 基于深度学习的器官自动分割研究进展[J]. 医疗卫生装备, 2020, 41(1): 85-94. DOI: 10.19745/j.1003-8868.2020021.
[2] 施俊, 汪琳琳, 王珊珊, 等. 深度学习在医学影像中的应用综述[J]. 中国图象图形学报, 2020, 25(10): 1953-1981. DOI: 10.11834/jig.200255.
[3] 李欣, 梁森, 黄正南, 等. 乳腺X线AI智能病灶检测[J]. 放射学实践, 2018, 33(10): 1029-1032. DOI: 10.13609/j.cnki.1000-0313.2018.10.010.
[4] KRIZHEVSKY A, SUTSKEVER I, HINTONG E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90. DOI: 10.1145/3065386.
[5] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[EB/OL]. (2015-04-10)[2023-11-25]. https://arxiv.org/abs/1409.1556. DOI: 10.48550/arXiv.1409.1556.
[6] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C] // 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2016: 770-778. DOI: 10.1109/cvpr.2016.90.
[7] SHELHAMER E,LONG J, DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(4): 640-651. DOI: 10.1109/TPAMI.2016.2572683.
[8] RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[C] // Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015: LNCS Volume 9351. Cham: Springer International Publishing Switzerland, 2015: 234-241. DOI: 10.1007/978-3-319-24574-4_28.
[9] ZHANG J W, ZHANG Y C, JIN Y Z, et al. MDU-Net: multi-scale densely connected U-Net for biomedical image segmentation[J]. Health Information Science and Systems, 2023, 11(1): 13. DOI: 10.1007/s13755-022-00204-9.
[10] ALOM M Z, HASAN M, YAKOPCIC C, et al. Recurrent residual convolutional neural network based on U-Net (R2U-Net) for medical image segmentation[EB/OL]. (2018-05-29)[2023-11-25]. https://arxiv.org/abs/1802.06955. DOI: 10.48550/arXiv.1802.06955.
[11] 彭璟, 罗浩宇, 赵淦森, 等. 深度学习下的医学影像分割算法综述[J]. 计算机工程与应用, 2021, 57(3): 44-57. DOI: 10.3778/j.issn.1002-8331.2010-0335.
[12] 曹玉红, 徐海, 刘荪傲, 等. 基于深度学习的医学影像分割研究综述[J]. 计算机应用, 2021, 41(8): 2273-2287. DOI: 10.11772/j.issn.1001-9081.2020101638.
[13] 周涛, 董雅丽, 霍兵强, 等. U-Net网络医学图像分割应用综述[J]. 中国图象图形学报, 2021, 26(9): 2058-2077. DOI: 10.11834/jig.200704.
[14] CHEN W L, ZHANG Y, HE J J, et al. Prostate segmentation using 2D bridged U-net[C] // 2019 International Joint Conference on Neural Networks (IJCNN). Piscataway, NJ: IEEE, 2019: 1-7. DOI: 10.1109/ijcnn.2019.8851908.
[15] 吴玲玉, 蓝洋, 夏海英. 基于卷积神经网络的眼底图像配准研究[J]. 广西师范大学学报(自然科学版), 2021, 39(5): 122-133. DOI: 10.16088/j.issn.1001-6600.2020122801.
[16] FU H Z, CHENG J, XU Y W, et al. Joint optic disc and cup segmentation based on multi-label deep network and polar transformation[J]. IEEE Transactions on Medical Imaging, 2018, 37(7): 1597-1605. DOI: 10.1109/tmi.2018.2791488.
[17] MILLETARI F, NAVAB N, AHMADI S A. V-Net: fully convolutional neural networks for volumetric medical image segmentation[C] // 2016 Fourth International Conference on 3D Vision (3DV). Los Alamitos, CA: IEEE Computer Society, 2016: 565-571. DOI: 10.1109/3dv.2016.79.
[18] ZHOU Z W, RAHMAN SIDDIQUEE M M, TAJBAKHSH N, et al. UNet++: a nested U-Net architecture for medical image segmentation[C] // Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support: LNCS Volume 11045. Cham: Springer, 2018: 3-11. DOI: 10.1007/978-3-030-00889-5_1.
[19] OKTAY O, SCHLEMPER J, FOLGOC L L, et al. Attention U-Net: learning where to look for the pancreas[EB/OL]. (2018-05-20)[2023-11-25]. https://arxiv.org/abs/1804.03999. DOI:10.48550/arXiv.1804.03999.
[20] 万黎明, 张小乾, 刘知贵, 等. 基于高效通道注意力的UNet肺结节CT图像分割[J]. 广西师范大学学报(自然科学版), 2022, 40(3): 66-75. DOI: 10.16088/j.issn.1001-6600.2021071202.
[21] 彭宝玉, 曹立佳. 多尺度融合注意力机制改进U-Net实现肺部感染区域分割[J]. 国外电子测量技术, 2023, 42(10): 177-183. DOI: 10.19652/j.cnki.femt.2305144.
[22] 谢飞, 权媚阳, 管子玉, 等. 基于深度学习的腹部多器官图像分割[J]. 西北大学学报(自然科学版), 2021, 51(1): 1-7. DOI: 10.16152/j.cnki.xdxbzr.2021-01-001.
[23] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C] // Advances in Neural Information Processing Systems 30 (NIPS 2017). Red Hook, NY: Curran Associates Inc., 2017: 6000-6010.
[24] DOSOVITSKIY A, BEYER L, KOLESNIKOVA, et al. An image is worth 16x16 words: transformers for image recognition at scale[EB/OL]. (2021-06-03)[2023-11-25]. https://arxiv.org/abs/2010.11929. DOI: 10.48550/arXiv.2010.11929.
[25] CHEN J N, LU Y Y, YU Q H, et al. TransUNet: transformers make strong encoders for medical image segmentation[EB/OL]. (2021-02-08)[2023-11-25]. https://arxiv.org/abs/2102.04306. DOI: 10.48550/arXiv.2102.04306.
[26] XIE E Z, WANG W H, YU Z D, et al. SegFormer: simple and efficient design for semantic segmentation with transformers[C] // Advances in Neural Information Processing Systems 34 (NeurIPS 2021). Red Hook, NY: Curran Associates Inc., 2021: 12077-12090.
[27] HUANG X H, DENG Z F, LI D D, et al. MISSFormer: an effective medical image segmentation transformer[EB/OL]. (2021-12-19)[2023-11-25]. https://arxiv.org/abs/2109.07162. DOI: 10.48550/arXiv.2109.07162.
[28] CAO H, WANG Y Y, CHEN J, et al. Swin-Unet: Unet-like pure transformer for medical image segmentation[C] // Computer Vision-ECCV 2022 Workshops: LNCS Volume 13803. Cham: Springer Nature Switzerland AG, 2023: 205-218. DOI: 10.1007/978-3-031-25066-8_9.
[29] WANG H Y, XIE S A, LIN L F, et al. Mixed transformer U-Net for medical image segmentation[C] // ICASSP 2022: 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Piscataway, NJ: IEEE, 2022: 2390-2394. DOI: 10.1109/icassp43922.2022.9746172.
[30] 李鸿, 邹俊颖, 谭茜成, 等. 面向医学图像分割的多注意力融合网络[J]. 计算机应用, 2022, 42(12): 3891-3899. DOI: 10.11772/j.issn.1001-9081.2021101737.
[31] 卢玲, 漆为民. 基于Transformer的脊椎CT图像分割[J]. 中国图象图形学报, 2023, 28(11): 3618-3628. DOI: 10.11834/jig.221084.
[32] Sage Bionetworks. Multi-atlas labeling beyond the cranial vault: workshop and challenge[EB/OL]. (2016-09-16)[2023-11-25]. https://www.synapse.org/Synapse:syn3193805/wiki/89480. DOI: 10.7303/syn3193805.
[33] BERNARD O, LALANDE A, ZOTTI C, et al. Deep learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: is the problem solved?[J]. IEEE Transactions on Medical Imaging, 2018, 37(11): 2514-2525. DOI: 10.1109/TMI.2018.2837502.
[34] LAMBERT Z, PETITJEAN C, DUBRAY B, et al. SegTHOR: segmentation of thoracic organs at risk in CT images[C] // 2020 Tenth International Conference on Image Processing Theory, Tools and Applications (IPTA). Piscataway, NJ: IEEE, 2020: 1-6. DOI: 10.1109/IPTA50016.2020.9286453.
[35] JI Y F, BAI H T, GE C J, et al. AMOS: a large-scale abdominal multi-organ benchmark for versatile medical image segmentation[C] // Advances in Neural Information Processing Systems 35 (NeurIPS 2022). Red Hook, NY: Curran Associates Inc., 2022: 36722-36732.
[36] HENDRYCKS D, GIMPEL K. Gaussian error linear units (GELUs)[EB/OL]. (2023-06-06)[2023-11-25]. https:// arxiv.org/abs/1606.08415. DOI: 10.48550/arXiv.1606.08415.
[37] ZHAO X Y, ZHANG P, SONG F, et al. Prior attention network for multi-lesion segmentation in medical images[J]. IEEE Transactions on Medical Imaging, 2022, 41(12): 3812-3823. DOI: 10.1109/tmi.2022.3197180.
[38] 李夏, 胡巍, 王子民. 基于Non-local的腰椎间盘突出患者多裂肌分割方法[J].桂林电子科技大学学报,2023,43(3): 217-222. DOI: 10.16725/j.cnki.cn45-1351/tn.2023.03.008.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed