COVID-Net CXR-2: An Enhanced Deep Convolutional Neural Network Design for Detection of COVID-19 Cases from Chest X-ray Images

Abstract

As the COVID-19 pandemic continues to devastate globally, the use of chest X-ray (CXR) imaging as a complimentary screening strategy to RT-PCR testing continues to grow given its routine clinical use for respiratory complaint. As part of the COVID-Net open source initiative, we introduce COVID-Net CXR-2, an enhanced deep convolutional neural network design for COVID-19 detection from CXR images built using a greater quantity and diversity of patients than the original COVID-Net. To facilitate this, we also introduce a new benchmark dataset composed of 19,203 CXR images from a multinational cohort of 16,656 patients from at least 51 countries, making it the largest, most diverse COVID-19 CXR dataset in open access form. The COVID-Net CXR-2 network achieves sensitivity and positive predictive value of 95.5%/97.0%, respectively, and was audited in a transparent and responsible manner. Explainability-driven performance validation was used during auditing to gain deeper insights in its decision-making behaviour and to ensure clinically relevant factors are leveraged for improving trust in its usage. Radiologist validation was also conducted, where select cases were reviewed and reported on by two board-certified radiologists with over 10 and 19 years of experience, respectively, and showed that the critical factors leveraged by COVID-Net CXR-2 are consistent with radiologist interpretations. While not a production-ready solution, we hope the open-source, open-access release of COVID-Net CXR-2 and the respective CXR benchmark dataset 1 will encourage researchers, clinical scientists, and citizen scientists to accelerate advancements and innovations in the fight against the pandemic.1 http://www.covid-net.ml Preprint. Under review.

MEDUSA: Multi-scale Encoder-Decoder Self-Attention Deep Neural Network Architecture for Medical Image Analysis

“…To evaluate the proposed model it is trained on the largest CXR dataset consisting of 19,203 CXR images [20]. The dataset [20] is constructed based on a cohort of 16,656 patients from at least 51 different countries.…”

“…To evaluate the proposed model it is trained on the largest CXR dataset consisting of 19,203 CXR images [20]. The dataset [20] is constructed based on a cohort of 16,656 patients from at least 51 different countries. There are total of 5,210 images from 2,815 SARS-CoV-2 positive patients and the rest of images are from 13,851 SARS-CoV-2 negative patients.…”

“…There are total of 5,210 images from 2,815 SARS-CoV-2 positive patients and the rest of images are from 13,851 SARS-CoV-2 negative patients. Interested readers can refer to [20] for more information on this dataset. Figure 5 demonstrates some examples of the CXR-2 dataset.…”

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

“…The model has reached an accuracy of 97.7%, 84.95%, and 97.03% on the mentioned datasets, respectively. Pavlova et al (2021) proposed the COVIDx8B dataset, the largest and most diverse COVID-19 CXR dataset in open access form, and the COVID-Net CXR-2 model, a CNN specially tailored for COVID-19 detection on CXR images using machine-driven design, which achieved an accuracy of 95.5%. Zhao et al (2021) used ResNet50V2 to classify the COVIDx8B dataset with an accuracy of 96.5% in the best scenario.…”

“…Therefore it is larger and more heterogeneous than earlier available datasets. However, there are only a few works that used this dataset so far (Pavlova et al, 2021;Zhao et al, 2021;Dominik, 2021). A recent survey on applications of artificial intelligence in COVID-19 pandemic (Khan et al, 2021) reviewed dozens of papers, including 16 papers on CNNs applied to CXR images and all of them used earlier available datasets which are smaller than COVIDx8B.…”