Detecting SARS-CoV-2 From Chest X-Ray Using Artificial Intelligence

Abstract

Chest radiographs (X-rays) combined with Deep Convolutional Neural Network (CNN) methods have been demonstrated to detect and diagnose the onset of COVID-19, the disease caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). However, questions remain regarding the accuracy of those methods as they are often challenged by limited datasets, performance legitimacy on imbalanced data, and have their results typically reported without proper confidence intervals. Considering the opportunity to address these issues, in this study, we propose and test six modified deep learning models, including VGG16, InceptionResNetV2, ResNet50, MobileNetV2, ResNet101, and VGG19 to detect SARS-CoV-2 infection from chest X-ray images. Results are evaluated in terms of accuracy, precision, recall, and f- score using a small and balanced dataset (Study One), and a larger and imbalanced dataset (Study Two). With 95% confidence interval, VGG16 and MobileNetV2 show that, on both datasets, the model could identify patients with COVID-19 symptoms with an accuracy of up to 100%. We also present a pilot test of VGG16 models on a multi-class dataset, showing promising results by achieving 91% accuracy in detecting COVID-19, normal, and Pneumonia patients. Furthermore, we demonstrated that poorly performing models in Study One (ResNet50 and ResNet101) had their accuracy rise from 70% to 93% once trained with the comparatively larger dataset of Study Two. Still, models like InceptionResNetV2 and VGG19's demonstrated an accuracy of 97% on both datasets, which posits the effectiveness of our proposed methods, ultimately presenting a reasonable and accessible alternative to identify patients with COVID-19.

Keywords: Artificial intelligence; COVID-19; SARS-CoV-2; chest X-ray; coronavirus; deep learning; imbalanced data; small data.

FIGURE 1.

Representative samples of chest X-ray images from the open source data repositories used in our proposed studies.

FIGURE 2.

Modified architecture with new classifier .

FIGURE 3.

VGG16 architecture used during this experiment.

FIGURE 4.

Study one confusion matrices for six different deep learning models applied on the test set.

FIGURE 5.

Training and validation accuracy throughout the execution of each model in study one.

FIGURE 6.

Training and validation loss throughout the execution of each model in study one.

FIGURE 7.

Study two confusion matrices for six different deep learning models applied on the test set.

FIGURE 8.

Training and validation accuracy throughout the execution of each model in study two.

FIGURE 9.

Training and validation loss throughout the execution of each model in study two.

FIGURE 10.

Heatmap of class activation on different layers.

FIGURE 11.

Model’s ability to identify important features on chest X-ray using VGG16.

FIGURE 12.

Model’s competency to identify essential features on chest X-ray using MobileNetV2.