Detection of COVID-19 lesions based on computed tomography using U-Net 2.5D and GAN

  • José Anatiel Landim UFPI
  • Edson Carvalho UFPI
  • João Otávio Diniz IFPI
  • Alcilene Sousa UFPI
  • Daniel Luz IFPI
  • Antônio Filho UFPI / IFPI


This paper proposes a computational method for automatically detecting suspected regions of COVID-19 from CT scans. COVID-19 has spread rapidly worldwide, infecting over 462 million people and causing over 6 million deaths. There are various methods to diagnose COVID-19, including imaging. The proposed method has five stages, including image acquisition, pre-processing, lung extraction, segmentation of suspected regions using U-Net 2.5D and Pix2Pix architectures, and result validation. The method achieved promising results, with 92% Dice for lung parenchyma segmentation, 82% Dice for suspected region segmentation using U-Net, and 71% Dice using Pix2Pix. It could potentially be integrated into clinical environments as a real aid system.


Bastos, M. L., Tavaziva, G., Abidi, S. K., Campbell, J. R., Haraoui, L.-P., Johnston, J. C., Lan, Z., Law, S., MacLean, E., Trajman, A., et al. (2020). Diagnostic accuracy of serological tests for covid-19: systematic review and meta-analysis. bmj, 370.

Costa, G., Paiva, A., Júnior, G. B., and Ferreira, M. (2021). Covid-19 automatic diagnosis with ct images using the novel transformer architecture. In Anais do XXI Simpósio Brasileiro de Computação Aplicada à Saúde, pages 293–301, Porto Alegre, RS, Brasil. SBC.

DE CARVALHO BRITO, V., DOS SANTOS, P. R. S., DE SALES CARVALHO, N. R., and DE CARVALHO FILHO, A. O. (2021). Covid-index: A texture-based approach to classifying lung lesions based on ct images. Pattern Recognition, 119:108083.

de Sousa Filho, E. L., Magno, P. G., Santos, C. J., da Silva, G. L., Diniz, J. O., and Quintanilha, D. B. (2022). Um método baseado em radiomics e mlp para diagnóstico automático de covid-19 a partir de raio-x de tórax. In Anais da X Escola Regional de Computação do Ceará, Maranhão e Piauí, pages 188–197. SBC.

Fang, C., Liu, Y., Liu, M., Qiu, X., Liu, Y., Li, Y., Wen, J., and Yang, Y. (2021). Unsupervised covid-19 lesion segmentation in ct using cycle consistent generative adversarial network.

Freire, N., Leão, P., Tiago, L., Gonçalves, A., Pinto, R., Santos, E., and Souto, E. (2022). Generalizability of cnn on predicting covid-19 from chest x-ray images. In Anais do XXII Simpósio Brasileiro de Computação Aplicada à Saúde, pages 36–47, Porto Alegre, RS, Brasil. SBC.

GOODFELLOW, I., POUGET-ABADIE, J., MIRZA, M., XU, B., WARDE-FARLEY, D., OZAIR, S., COURVILLE, A., and BENGIO, Y. (2014). Generative adversarial nets. in: Advances in neural information processing systems. page 2072–2680.

Hasanzadeh, N., Paima, S. S., Bashirgonbadi, A., Naghibi, M., and Soltanian-Zadeh, H. (2020). Segmentation of covid-19 infections on ct: Comparison of four unet-based networks. In 2020 27th National and 5th International Iranian Conference on Biomedical Engineering (ICBME), pages 222–225.

Isola, P., Zhu, J.-Y., Zhou, T., and Efros, A. A. (2017). Image-to-image translation with conditional adversarial networks. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

JHU, J. H. U. (2021). Coronavirus covid-19. Accessed Jul. 2021].

Jun, M., Cheng, G., Yixin, W., Xingle, A., Jiantao, G., Ziqi, Y., Minqing, Z., Xin, L., Xueyuan, D., Shucheng, C., Hao, W., Sen, M., Xiaoyu, Y., Ziwei, N., Chen, L., Lu, T., Yuntao, Z., Qiongjie, Z., Guoqiang, D., and Jian, H. (2020). Covid-19 ct lung and infection segmentation dataset.

MedSeg (2020). Covid-19 ct segmentation dataset.

Müller, D., Soto Rey, I., Kramer, F., and . (2020). Automated chest ct image segmentation of covid-19 lung infection based on 3d u-net.

Oulefki, A., Agaian, S., Trongtirakul, T., and Kassah Laouar, A. (2021). Automatic covid-19 lung infected region segmentation and measurement using ct-scans images. Pattern Recognition, 114:107747.

Ouyang, X., Huo, J., Xia, L., Shan, F., Liu, J., Mo, Z., Yan, F., Ding, Z., Yang, Q., Song, B., Shi, F., Yuan, H., Wei, Y., Cao, X., Gao, Y., Wu, D., Wang, Q., and Shen, D. (2020). Dual-sampling attention network for diagnosis of covid-19 from community acquired pneumonia. IEEE Transactions on Medical Imaging, 39(8):2595–2605.

Ramos, R., Ferreira, C., and Simão, A. (2022). The survival rate among unvaccinated, first dose, and second dose brazilian hospitalized and icu covid patients by age group. In Anais do XXII Simpósio Brasileiro de Computação Aplicada à Saúde, pages 48–59, Porto Alegre, RS, Brasil. SBC.

Ronneberger, O., Fischer, P., and Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, page 234–241.

Silva, L. d. J., Cortes, O., and Diniz, J. (2023). A novel ensemble cnn model for covid-19 classification in computerized tomography scans. Results in Control and Optimization, page 100215.

Silva, L. S., Diniz, J. O., da Silva, G. L., Silva, A. C., and Paiva, A. C. (2021). Segmentation and reconstruction of the lung in ct affected by covid-19 using deep learning and adaptive convex hull. In 17th International Symposium on Medical Information Processing and Analysis, volume 12088, pages 56–64. SPIE.

Vu, M. H., Grimbergen, G., Nyholm, T., and Löfstedt, T. (2020). Evaluation of multislice inputs to convolutional neural networks for medical image segmentation. Medical Physics, 47(12):6216–6231.

Wang, G., Liu, X., Li, C., Xu, Z., Ruan, J., Zhu, H., Meng, T., Li, K., Huang, N., and Zhang, S. (2020). A noise-robust framework for automatic segmentation of covid-19 pneumonia lesions from ct images. IEEE Transactions on Medical Imaging, 39(8):2653–2663.

Xu, Z., Cao, Y., Jin, C., Shao, G., Liu, X., Zhou, J., Shi, H., and Feng, J. (2020). Gasnet: Weakly-supervised framework for covid-19 lesion segmentation.
Como Citar

Selecione um Formato
LANDIM, José Anatiel; CARVALHO, Edson; DINIZ, João Otávio; SOUSA, Alcilene; LUZ, Daniel; FILHO, Antônio. Detection of COVID-19 lesions based on computed tomography using U-Net 2.5D and GAN. In: SIMPÓSIO BRASILEIRO DE COMPUTAÇÃO APLICADA À SAÚDE (SBCAS), 23. , 2023, São Paulo/SP. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2023 . p. 83-94. ISSN 2763-8952. DOI: