Dataset Information

Classification of X-ray images into COVID-19, pneumonia, and TB using cGAN and fine-tuned deep transfer learning models

ABSTRACT:

Purpose

The rapid increase in the spread of the coronavirus disease of 2019 (COVID-19) has led to a need for reliable, effective, and readily available testing on a large scale. While diagnostic testing has been a support to public health, newer technology can be used to provide low-cost and convenient test options for patients. X-ray scanning can be performed to resolve this issue and produce quicker and more precise results. Currently, a radiologist is required to examine these X-ray images. However, deep convolutional neural networks can also be used to perform X-ray examinations and employed for the detection of COVID-19. We propose a Conditional Generative Adversarial Network (cGAN) with a fine-tuned deep transfer learning model to classify chest X-rays into six categories: COVID-Mild, COVID-Medium, COVID-Severe, Normal, Pneumonia, and Tuberculosis.

Methods

A total of 1229 images were taken to form a dataset containing six classes corresponding to the six categories. A cGAN was used to increase the number of images. Generative Adversarial Networks (GAN) are used to train a model for generating new images. cGAN is an extension of GAN consisting of a generator and discriminator network that are trained simultaneously to optimize the model. The generated images were then trained using deep transfer learning models such as ResNet50, Xception, and DenseNet-169 to achieve the classification into six classes.

Results

The proposed model helped achieve a training and validation accuracy of up to 98.20 % and 94.21 % respectively. The model was able to achieve a test accuracy of 93.67 %. The use of cGAN not only helped to increase the size of the training dataset but it also helped to reduce the problem of over-fitting.

Conclusion

The proposed approach will help to diagnose COVID-19 quickly at an early stage.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42600-021-00174-z.

SUBMITTER: Mehta T

PROVIDER: S-EPMC8494631 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:The COVID-19 pandemic is causing a major outbreak in more than 150 countries around the world, having a severe impact on the health and life of many people globally. One of the crucial step in fighting COVID-19 is the ability to detect the infected patients early enough, and put them under special care. Detecting this disease from radiography and radiology images is perhaps one of the fastest ways to diagnose the patients. Some of the early studies showed specific abnormalities in the chest radiograms of patients infected with COVID-19. Inspired by earlier works, we study the application of deep learning models to detect COVID-19 patients from their chest radiography images. We first prepare a dataset of 5000 Chest X-rays from the publicly available datasets. Images exhibiting COVID-19 disease presence were identified by board-certified radiologist. Transfer learning on a subset of 2000 radiograms was used to train four popular convolutional neural networks, including ResNet18, ResNet50, SqueezeNet, and DenseNet-121, to identify COVID-19 disease in the analyzed chest X-ray images. We evaluated these models on the remaining 3000 images, and most of these networks achieved a sensitivity rate of 98% ( ± 3%), while having a specificity rate of around 90%. Besides sensitivity and specificity rates, we also present the receiver operating characteristic (ROC) curve, precision-recall curve, average prediction, and confusion matrix of each model. We also used a technique to generate heatmaps of lung regions potentially infected by COVID-19 and show that the generated heatmaps contain most of the infected areas annotated by our board certified radiologist. While the achieved performance is very encouraging, further analysis is required on a larger set of COVID-19 images, to have a more reliable estimation of accuracy rates. The dataset, model implementations (in PyTorch), and evaluations, are all made publicly available for research community at https://github.com/shervinmin/DeepCovid.git.