Location:Home / Journals / Article Detail

Advances in Computer and Communication

DOI:http://dx.doi.org/10.26855/acc.2022.06.002

PDF Download

Classification of COVID-19 Chest Radiographs Based on Convolutional Neural Network

Ziheng Liu, Zhi Li*

College of Electronics and Information Engineering, Sichuan University, Chengdu, Sichuan, China.

*Corresponding author: Zhi Li

Date: June 27,2022 Hits: 296

Abstract

As one of the most lethal infectious diseases in the world today, the prevention, control and treatment of COVID-19 have become the focus of global public health. In view of the low efficiency of manual detection methods for COVID-19 chest radiographs, and the possibility of misdiagnosis and missed diagnosis, a DA-COVID Net model for image classification of COVID-19 chest radiographs was proposed. Based on the residual network ResNet50 model, the parallel loca-tion attention module and channel attention module are introduced to enhance the feature representation, and then the classification output is performed after fusion. After pre-processing such as clipping and data enhancement, the data set was put into DA-COVID Net model for training. The experimental results show that DA-COVID Net has achieved 97.7% accuracy in the classification of COVID-19 chest radiographs, which is significantly improved compared with other models. With excellent performance evaluation indexes and fast convergence, DA-COVID Net can provide convenient and reliable basis for clinical diagnosis of COVID-19.

References

[1] Xu, X., Yu, C., Zhang, L., et al. (2020). Imaging features of 2019 novel coronavirus pneumonia [J]. European journal of nuclear medicine and molecular imaging, 2020, 47(5): 1022-1023.

[2] Huang, Y., Liu, A., Liang, L., et al. (2018). Diagnostic value of blood parameters for community-acquired pneumonia [J]. International Immunopharmacology, 2018, 64: 10-15.

[3] National Health Commission. COVID-19 Diagnosis and Treatment Protocol (Trial Version 8) [EB/OL]. 2020. http://www.nhc.gov.cn/yzygj/s7653p/202008/0a7bdf12bd4b46e5bd28ca7f9a7f5e5a/files/a449a3e2e2c94d9a856d5faea2ff0f94.pdf.

[4] Xie, X., Zhong, Z., Zhao, W., et al. (2020). Chest CT for Typical 2019-n CoV Pneumonia: Relationship to Negative RT-PCR Testing [J]. Radiology, 2020, 296(2): 41-45.

[5] K. He, X. Zhang, S. Ren, and J. Sun. (2016). Deep residual learning for image recognition, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016, pp. 770-778.

[6] Wang, X., Girshick, R., Gupta, A., et al. (2018). Non-local neural networks [C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2018: 7794-7803.

[7] Jie, H., Li, S., Gang, S., et al. (2017). Squeeze-and-Excitation Networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 42(8): 2011-2023.

[8] Liang-Chieh Chen, George Papandreou, Iasonas Kokkinos, Kevin Murphy, and Alan L. Yuille. (2018). Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs. IEEE Transactions on Pattern Analysis and Machine Intelligence, 40(4): 834-848, 2018.

[9] Yu, F., Koltun, V., and Funkhouser, T. (2017). Dilated residual networks [C]//Proceedings of the IEEE conference on computer vision and pattern recognition, 2017: 472-480.

Full-Text HTML

Classification of COVID-19 Chest Radiographs Based on Convolutional Neural Network

How to cite this paper: Ziheng Liu, Zhi Li. (2022) Classification of COVID-19 Chest Radiographs Based on Convolutional Neural Network. Advances in Computer and Communication3(1), 23-28.

DOI: http://dx.doi.org/10.26855/acc.2022.06.002