ConvUCaps：基于卷积胶囊网络的医学图像分割模型

doi:10.3778/j.issn.1002-8331.2302-0002

摘要/Abstract

摘要： 在医学影像分割领域，U-Net网络是目前最成功和最受关注的方法之一，但是U-Net本质上是一种经过改造的全卷积神经网络模型，要获得更为全面和准确的局部-整体关系，不但需要增加网络层次从而加大计算量，而且效果也并不明显。胶囊网络提供了一种有效的建模图像的局部与整体关系的方法，可以用更少的参数取得好的性能。但原始的胶囊网络并没有充分考虑图像局部特征的粒度问题，将其应用在医学图像分割领域还需进一步改造。因此，提出一种将U-Net和胶囊网络相结合的医学图像分割模型ConvUCaps。该模型对U-Net的编码器部分进行改进，使用卷积模块学习不同尺度的局部特征，然后通过胶囊模块学习高层特征，并建模局部与整体之间的关系。实验结果表明，相比U-Net、UNet++、SegCaps、Matwo-CapsNet网络，ConvUCaps提高了分割精度和收敛速度，同时，与单纯基于胶囊网络的分割模型相比，显著减少了推理时间。

关键词: 医学图像分割, 卷积神经网络, U-Net网络, 胶囊网络

Abstract: In the field of medical image segmentation, U-Net network is one of the most successful and concerned methods at present. However, U-Net is essentially a modified fully convolutional neural network model. To obtain a more comprehensive and accurate local-whole relationship, it is not only necessary to increase the network level which increases the amount of calculation, but also the effect is not obvious. Capsule network provides an effective method to model the local and whole relationship of images, which can achieve good performance with fewer parameters. However, the original capsule network does not fully consider the granularity of local features of the image, and its application in the field of medical image segmentation needs further improvement. Therefore, this paper proposes a medical image segmentation model named ConvUCaps, which combines U-Net and capsule network. This model improves the encoder part of U-Net, uses convolutional module to learn local features of different scales, and then uses capsule module to learn high-level features and model the local-whole relationship. The experimental results show that compared with U-Net, UNet++, SegCaps and Matwo-CapsNet networks, ConvUCaps improves the segmentation accuracy and convergence speed. At the same time, the inference time is significantly reduced compared with the segmentation model based solely on capsule network.

Key words: medical image segmentation, convolutional neural network, U-Net network, capsule network

邓希泉, 陈刚. ConvUCaps：基于卷积胶囊网络的医学图像分割模型[J]. 计算机工程与应用, 2024, 60(8): 258-266.

DENG Xiquan, CHEN Gang. ConvUCaps: Medical Image Segmentation Model Based on Convolutional Capsule Network[J]. Computer Engineering and Applications, 2024, 60(8): 258-266.

参考文献

[1] 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.
PENG J, LUO H Y, ZHAO G S, et al. Survey of medical image segmentation algorithm in deep learning[J]. Computer Engineering and Applications, 2021, 57(3): 44-57.
[2] ILHAN U, ILHAN A. Brain tumor segmentation based on a new threshold approach[J]. Procedia Computer Science, 2017, 120: 580-587.
[3] PRABIN A, VEERAPPAN J. Automatic segmentation of lung CT images by CC based region growing[J]. Journal of Theoretical and Applied Information Technology, 2014, 68(1): 63-69.
[4] MAHMOOD N, SHAH A, WAQAS A, et al. Image segmentation methods and edge detection: an application to knee joint articular cartilage edge detection[J]. Journal of Theoretical and Applied Information Technology, 2015, 71(1): 87-96.
[5] WANG R, LEI T, CUI R, et al. Medical image segmentation using deep learning: a survey[J]. IET Image Processing, 2022, 16(5): 1243-1267.
[6] RONNEBERGER O, FISCHER P, BROX T. U-net: convolutional networks for biomedical image segmentation[C]//Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, 2015: 234-241.
[7] LE N, YAMAZAKI K, QUACH K G, et al. A multi-task contextual atrous residual network for brain tumor detection & segmentation[C]//Proceedings of the 2020 25th International Conference on Pattern Recognition, 2021: 5943-5950.
[8] FAN T, WANG G, LI Y, et al. MA-NET: a multi-scale attention network for liver and tumor segmentation[J]. IEEE Access, 2020, 8: 179656-179665.
[9] FAN T, WANG G, WANG X, et al. MSN-Net: a multi-scale context nested U-Net for liver segmentation[J]. Signal, Image and Video Processing, 2021, 15: 1089-1097.
[10] GUO C, SZEMENYEI M, YI Y, et al. Sa-unet: Spatial attention u-net for retinal vessel segmentation[C]//Proceedings of the 2020 25th International Conference on Pattern Recognition, 2021: 1236-1242.
[11] YAHYATABAR M, JOUVET P, CHERIET F. Dense-Unet: a light model for lung fields segmentation in Chest X-Ray images[C]//Proceedings of the 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society, 2020: 1242-1245.
[12] SABOUR S, FROSST N, HINTON G E. Dynamic routing between capsules[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems, 2017: 3859-3869.
[13] LALONDE R, BAGCI U. Capsules for object segmentation[J]. arXiv:1804.04241, 2018.
[14] LALONDE R, XU Z, IRMAKCI I, et al. Capsules for biomedical image segmentation[J]. Medical Image Analysis, 2021, 68:101889.
[15] BONHEUR S, ?TERN D, PAYER C, et al. Matwo-CapsNet: a multi-label semantic segmentation capsules network[C]//Proceedings of the International Conference on Medical Image Computing and Computer-Assisted intervention, 2019: 664-672.
[16] LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015: 3431-3440.
[17] ZHANG J, JIN Y, XU J, et al. MDU-Net: multi-scale densely connected U?Net for biomedical image segmentation[J]. arXiv:1812.00352, 2018.
[18] OKTAY O, SCHLEMPER J, FOLGOC L L, et al. Attention U-Net: learning where to look for the pancreas[J]. arXiv:1804.03999, 2018.
[19] ZHOU Z, RAHMAN SIDDIQUEE M M, TAJBAKHSH N, et al. UNet++: a nested U-Net architecture for medical image segmentation[C]//Proceedings of the Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support, 2018: 3-11.
[20] ISENSEE F, JAEGER P F, KOHL S A A, et al. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation[J]. Nature Methods, 2021, 18(2): 203-211.
[21] MOBINY A, NGUYEN H V. Fast CapsNet for lung cancer screening[C]//Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, 2018: 741-749.
[22] HE Y, QIN W, WU Y, et al. Automatic left ventricle segmentation from cardiac magnetic resonance images using a capsule network[J]. Journal of X-ray Science and Technology, 2020, 28(3): 541-553.
[23] JIMéNEZ-SáNCHEZ A, ALBARQOUNI S, MATEUS D. Capsule networks against medical imaging data challenges[C]//Proceedings of the Intravascular Imaging and Computer Assisted Stenting and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis, 2018: 150-160.
[24] SHIRAISHI J, KATSURAGAWA S, IKEZOE J, et al. Development of a digital image database for chest radiographs with and without a lung nodule: receiver operating characteristic analysis of radiologists' detection of pulmonary nodules[J].American Journal of Roentgenology, 2000, 174(1): 71-74.
[25] 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.
[26] GINNEKEN B V, STEGMANN M B, LOOG M. Segmentation of anatomical structures in chest radiographs using supervised methods: a comparative study on a public database[J]. Medical Image Analysis, 2006, 10(1): 19-40.