Education, Science, Technology, Innovation and Life
Open Access
Sign In

CT Image Segmentation Network of COVID-19 Based on Multi-scale Attention and Probability Preserving Pooling

Download as PDF

DOI: 10.23977/jipta.2023.060105 | Downloads: 10 | Views: 430

Author(s)

Zhiwen Yang 1,2, Mingju Chen 1,2

Affiliation(s)

1 School of Automation and Information Engineering, Sichuan University of Science and Engineering, Zigong, 643000, China
2 Artificial Intelligence Key Laboratory of Sichuan Province, Yibin, 644000, China

Corresponding Author

Mingju Chen

ABSTRACT

In order to effectively improve the accuracy of U-Net segmentation of COVID-19 CT images, a novel COVID-19 CT image segmentation model based on multi-scale attention and class probability Preserving pooling auxiliary classifier was proposed. In view of the shortcomings of U-Net in using single-size standard two-dimensional convolution to extract image feature information, a multi-scale attention module is adopted to enhance the performance of U-Net in extracting multi-scale information by using three groups of attention modules mixed with different sizes of deep strip convolution and spatial domain channel domain. Aiming at the disadvantage of information loss when using maximum pooling in the U-Net lower sampling layer, two-dimensional convolution is used to reduce the size of the feature map, and an auxiliary classifier is used to calculate the coarse segmentation semantic probability map of each layer of the encoder and the loss is calculated by using the GT tag map of the class probability Preserving pooling to optimize the classification performance of the network. The network model is tested for image segmentation in the standard COVID-19 CT image dataset, and the segmentation accuracy for the infected area is 87.22%, the recall rate is 84.55%, and the intersection/merger ratio is 75.98%.

KEYWORDS

COVID-19, deep learning, image segmentation, multiscale attention, class probability preserving pooling

CITE THIS PAPER

Zhiwen Yang, Mingju Chen, CT Image Segmentation Network of COVID-19 Based on Multi-scale Attention and Probability Preserving Pooling. Journal of Image Processing Theory and Applications (2023) Vol. 6: 54-61. DOI: http://dx.doi.org/10.23977/jipta.2023.060105.

REFERENCES

[1] Shelhamer, Evan et al. "Fully convolutional networks for semantic segmentation." 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2014): 3431-3440.
[2] Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation//International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, Cham, 2015: 234-241.
[3] Takikawa, Towaki et al. "Gated-SCNN: Gated Shape CNNs for Semantic Segmentation." 2019 IEEE/CVF International Conference on Computer Vision (ICCV) (2019): 5228-5237.
[4] Fan D P, Zhou T, Ji G P, et al. Inf-Net: Automatic COVID-19 Lung Infection Segmentation from CT Images. IEEE Transactions on Medical Imaging, 2020, 39(8): 2626 – 2637.
[5] Zhu Lanyun et al. "Learning Statistical Texture for Semantic Segmentation." 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2021): 12532-12541.
[6] Woo S, Park J, Lee J Y, et al. Cbam: Convolutional block attention module//Proceedings of the European conference on computer vision (ECCV). 2018: 3-19.
[7] Szegedy, Christian et al. "Going deeper with convolutions." 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2014): 1-9.
[8] Sun Ke et al. "Deep High-Resolution Representation Learning for Human Pose Estimation." 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2019): 5686-5696.
[9] Guo M H, Lu C Z, Hou Q, et al. Segnext: Rethinking convolutional attention design for semantic segmentation. arXiv preprint arXiv:2209.08575, 2022..
[10] Peng Chao et al. "Large Kernel Matters-Improve Semantic Segmentation by Global Convolutional Network." 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2017): 1743-1751.
[11] Zhao Hengshuang et al. "Pyramid Scene Parsing Network." 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2016): 6230-6239.
[12] Han D, Yoo J, Oh D. See Through Net: Resurrection of Auxiliary Loss by Preserving Class Probability Information//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2022: 4463-4472.
[13] Ma J, Wang Y, An X, et al. Towards Efficient COVID-19 CT Annotation: A Benchmark for Lung and Infection Segmentation. arXiv, 2020.
[14] Morozov S P, Andreychenko A E, Pavlov N A, et al. MosMedData: Chest CT Scans with COVID-19 Related Findings Dataset. 2020.
[15] Shelhamer Evan et al. "Fully convolutional networks for semantic segmentation." 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2014): 3431-3440.
[16] He Kaiming et al. "Mask R-CNN." 2017 IEEE International Conference on Computer Vision (ICCV) (2017): 2980-2988.

Downloads: 1117
Visits: 97653

Sponsors, Associates, and Links


All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.