This is a preprint, a preliminary version of a manuscript that has not completed peer review at a journal. Research Square does not conduct peer review prior to posting preprints. The posting of a preprint on this server should not be interpreted as an endorsement of its validity or suitability for dissemination as established information or for guiding clinical practice.
Research Article
Tahmina Zebin
University of East Anglia
Corresponding Author
ORCiD: https://orcid.org/0000-0003-0437-0570
License:
This work is licensed under a CC BY 4.0 License. Read Full License
Chest X-rays are playing an important role in the testing and diagnosis of COVID-19 disease in the recent pandemic. However, due to the limited amount of labelled medical images, automated classification of these images for positive and negative cases remains the biggest challenge in their reliable use in diagnosis and disease progression. We applied and implemented a transfer learning pipeline for classifying COVID-19 chest X-ray images from two publicly available chest X-ray datasets {https://github.com/ieee8023/covid-chestxray-dataset},{https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia}}. The classifier effectively distinguishes inflammation in lungs due to COVID-19 and pneumonia (viral and bacterial) from the ones with no infection (normal). We have used multiple pre-trained convolutional backbones as the feature extractor and achieved an overall detection accuracy of 91.2% , 95.3%, 96.7% for the VGG16, ResNet50 and EfficientNetB0 backbones respectively. Additionally, we trained a generative adversarial framework (a cycleGAN) to generate and augment the minority COVID-19 class in our approach. For visual explanations and interpretation purposes, we visualized the regions of input that are important for predictions and a gradient class activation mapping (Grad-CAM) technique is used in the pipeline to produce a coarse localization map of the highlighted regions in the image. This activation map can be used to monitor affected lung regions during disease progression and severity stages.
Keywords
Deep learning, Covid19, chest X-ray
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
This preprint is available for download as a PDF.
Loading...
Badges
Prescreen
This manuscript passed our Prescreen assessment
Complete author information
Appropriate declaration statements
Potential risks to human health
Prescreen
History
CURRENT STATUS: Posted
View the published article at Applied Intelligence.
Version 1
Posted 15 Jun, 2020
No community comments so far
Metrics
Comments: 0
PDF Downloads: ...
HTML Views: ...
Learn more about our company.
This is a preprint. It has not completed peer review.
Research Article
Tahmina Zebin
University of East Anglia
Corresponding Author
ORCiD: https://orcid.org/0000-0003-0437-0570
Badges
Prescreen
This manuscript passed our Prescreen assessment
Complete author information
Appropriate declaration statements
Potential risks to human health
Prescreen
History
CURRENT STATUS: Posted
View the published article at Applied Intelligence.
Version 1
Posted 15 Jun, 2020
No community comments so far
Metrics
Comments: 0
PDF Downloads: ...
HTML Views: ...
License:
This work is licensed under a CC BY 4.0 License. Read Full License
View the published article at Applied Intelligence.
Chest X-rays are playing an important role in the testing and diagnosis of COVID-19 disease in the recent pandemic. However, due to the limited amount of labelled medical images, automated classification of these images for positive and negative cases remains the biggest challenge in their reliable use in diagnosis and disease progression. We applied and implemented a transfer learning pipeline for classifying COVID-19 chest X-ray images from two publicly available chest X-ray datasets {https://github.com/ieee8023/covid-chestxray-dataset},{https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia}}. The classifier effectively distinguishes inflammation in lungs due to COVID-19 and pneumonia (viral and bacterial) from the ones with no infection (normal). We have used multiple pre-trained convolutional backbones as the feature extractor and achieved an overall detection accuracy of 91.2% , 95.3%, 96.7% for the VGG16, ResNet50 and EfficientNetB0 backbones respectively. Additionally, we trained a generative adversarial framework (a cycleGAN) to generate and augment the minority COVID-19 class in our approach. For visual explanations and interpretation purposes, we visualized the regions of input that are important for predictions and a gradient class activation mapping (Grad-CAM) technique is used in the pipeline to produce a coarse localization map of the highlighted regions in the image. This activation map can be used to monitor affected lung regions during disease progression and severity stages.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
This preprint is available for download as a PDF.
Loading...