COVID-19 Detection on Chest X-Ray Images: A comparison of CNN architectures and ensembles

TL;DR

This study compares 21 CNN architectures and ensembles for COVID-19 detection in chest X-ray images, demonstrating that ensemble methods significantly improve accuracy and outperform recent results on a large international dataset.

Contribution

It provides a comprehensive comparison of multiple CNN architectures and shows the effectiveness of ensemble models for COVID-19 detection in CXR images.

Findings

DenseNet169 achieved 98.15% accuracy as a standalone model.

Ensemble of five DenseNet169 models reached 99.25% accuracy.

Ensembles outperform individual CNNs and recent benchmarks.

Abstract

COVID-19 quickly became a global pandemic after only four months of its first detection. It is crucial to detect this disease as soon as possible to decrease its spread. The use of chest X-ray (CXR) images became an effective screening strategy, complementary to the reverse transcription-polymerase chain reaction (RT-PCR). Convolutional neural networks (CNNs) are often used for automatic image classification and they can be very useful in CXR diagnostics. In this paper, 21 different CNN architectures are tested and compared in the task of identifying COVID-19 in CXR images. They were applied to the COVIDx8B dataset, a large COVID-19 dataset with 16,352 CXR images coming from patients of at least 51 countries. Ensembles of CNNs were also employed and they showed better efficacy than individual instances. The best individual CNN instance results were achieved by DenseNet169, with an accuracy of 98.15% and an F1 score of 98.12%. These were further increased to 99.25% and 99.24%, respectively, through an ensemble with five instances of DenseNet169. These results are higher than those obtained in recent works using the same dataset.