Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Aug:135:104588.
doi: 10.1016/j.compbiomed.2021.104588. Epub 2021 Jun 22.

COVID-19 deep classification network based on convolution and deconvolution local enhancement

Lingling Fang  1 , Xin Wang  2
Affiliations

COVID-19 deep classification network based on convolution and deconvolution local enhancement

Lingling Fang et al. Comput Biol Med. 2021 Aug.

Abstract

Computer Tomography (CT) detection can effectively overcome the problems of traditional detection of Corona Virus Disease 2019 (COVID-19), such as lagging detection results and wrong diagnosis results, which lead to the increase of disease infection rate and prevalence rate. The novel coronavirus pneumonia is a significant difference between the positive and negative patients with asymptomatic infections. To effectively improve the accuracy of doctors' manual judgment of positive and negative COVID-19, this paper proposes a deep classification network model of the novel coronavirus pneumonia based on convolution and deconvolution local enhancement. Through convolution and deconvolution operation, the contrast between the local lesion region and the abdominal cavity of COVID-19 is enhanced. Besides, the middle-level features that can effectively distinguish the image types are obtained. By transforming the novel coronavirus detection problem into the region of interest (ROI) feature classification problem, it can effectively determine whether the feature vector in each feature channel contains the image features of COVID-19. This paper uses an open-source COVID-CT dataset provided by Petuum researchers from the University of California, San Diego, which is collected from 143 novel coronavirus pneumonia patients and the corresponding features are preserved. The complete dataset (including original image and enhanced image) contains 1460 images. Among them, 1022 (70%) and 438 (30%) are used to train and test the performance of the proposed model, respectively. The proposed model verifies the classification precision in different convolution layers and learning rates. Besides, it is compared with most state-of-the-art models. It is found that the proposed algorithm has good classification performance. The corresponding sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and precision are 0.98, 0.96, 0.98, and 0.97, respectively.

Keywords: COVID-19 classification network; Convolution; Deconvolution; Enhanced CT features; ROI extraction.

PubMed Disclaimer

Conflict of interest statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Figures

Image 1
Graphical abstract
Fig. 1
Fig. 1
The COVID-19 CT images from multi-angles. (a) The process of COVID-19 infection; (b) The CT image from multi-angles.
Fig. 2
Fig. 2
The CT image of COVID-19 patients at different stages. (a) Ground glass shadow; (b) Fireworks like ground glass shadow; (c) Consolidation shadow; (d) White lung.
Fig. 3
Fig. 3
Comparison of image features between the AIDS and COVID-19 diseases. (a) Image feature of AIDS; (b) Image feature of COVID-19; (c) Image feature of the coexistence of AIDS and COVID-19.
Fig. 4
Fig. 4
The coverage ratio of the effective distribution infected with COVID-19. (a) Left lung; (b) Right lung.
Fig. 5
Fig. 5
The ROI extraction. (a) The ROI of left lung; (b) The extracted region of the lung CT image; (c) The ROI of the right lung.
Fig. 6
Fig. 6
The process chart of the contrast enhancement.
Fig. 7
Fig. 7
The contrast enhancement of the lesion region. (a) The original image; (b) The ROI enhanced image.
Fig. 8
Fig. 8
The structure of the CNN model.
Fig. 9
Fig. 9
The visualization of the COVID-19 feature vector. (a) The COVID-19 positive image; (b) The COVID-19 negative image. (The first and the second columns represent the original image and the corresponding visual feature vector distribution, respectively.)
Fig. 10
Fig. 10
Deep convolution network model.
Fig. 11
Fig. 11
The feature map with convolution kernels of different sizes.
Fig. 12
Fig. 12
The kernel image in different convolution layers. (a) Convolution kernels of different sizes; (b) Characteristic map of fully connected layers.
Fig. 13
Fig. 13
The COVID-CT dataset. (a) COVID-19 positive image; (b) COVID-19 negative image.
Fig. 14
Fig. 14
The training process of the deep convolution model. (a) The learning rate and loss curve of the proposed deep convolution model in the training process; (b) The corresponding parameters of the proposed model using different iterations.
Fig. 15
Fig. 15
Index values under different learning rates.
Fig. 16
Fig. 16
The confusion matrix of the proposed deep convolution network.
Fig. 17
Fig. 17
The classification precision under the different datasets. (a) The precision of the model test under the different datasets; (b) The time required to test different datasets.
Fig. 18
Fig. 18
The result of the different convolution layers. (a) The Classification precision of the different convolution layers; (b) The classification precision; (c) The loss rate.
Fig. 19
Fig. 19
The classification precision under different proportions of the training dataset.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Kraemer Moritz U.G., Scarpino Samuel V., Vukosi Marivate, et al. Data curation during a pandemic and lessons learned from COVID-19. Nat. Comput. Sci. Jan. 2021;1(1):9–10. - PubMed
    1. Harsh Panwar, Gupta P.K. Siddiqui Mohammad Khubeb, et all, “A Deep Learning and Grad-CAM based Color Visualization Approach for Fast Detection of COVID-19 Cases using Chest X-ray and CT-Scan Images,” Chaos. Solitons Fractals. Aug. 2020;140 doi: 10.1016/j.chaos.2020.110190. - DOI - PMC - PubMed
    1. Ismael Aras M., Abdulkadir Şengür. Deep learning approaches for COVID-19 detection based on chest X-ray images. Expert Syst. Appl. Jan. 2021;164 doi: 10.1016/J.ESWA.2020.114054. - DOI - PMC - PubMed
    1. Abdulhadee Yakoh, Umaporn Pimpitak, Sirirat Rengpipat, et al. ” Biosensors and Bioelectronics; Dec. 2020. Paper-based Electrochemical Biosensor for Diagnosing COVID-19: Detection of SARS-CoV-2 Antibodies and Antigen. - DOI - PMC - PubMed
    1. Rai Praveen, Krishna Kumar Ballamoole, Kumar Deekshit Vijaya, et al. Detection technologies and recent developments in the diagnosis of COVID-19 infection. Appl. Microbiol. Biotechnol. Jan. 2021:1–15. doi: 10.1007/S00253-020-11061-5. - DOI - PMC - PubMed

Publication types