STCovidNet: Automatic Detection Model of Novel Coronavirus Pneumonia Based on Swin Transformer

Abstract

The novel coronavirus disease 2019 (COVID-19) has emerged as an enormous challenge facing China today. Preventive Medicine physicians and Artificial Intelligence (AI) researchers try to improve the ability to early automatic warning of coronavirus infections, promote epidemic prevention, and reduce medical costs using deep learning methods. In this work, we build an extensive database of chest computed tomography (CT) scans with image data from domestic and international open-source medical datasets. Swin Transformer is chosen as the backbone network to establish a model (STCovidNet) for the prediction of COVID-19. We then compare the performance of our technique against that of Vision Transformer (ViT) and Convolutional Neural Network (CNN). Next, to visualize our model's high-dimensional outputs in 2-dimensional space, we apply t-distributed stochastic neighbor embedding (t-SNE) as the dimension-reduction strategy. Finally, we employ gradient-weighted class activation mapping (Grad-CAM) to present a class activation map. The results indicate that STCovidNet’s performance surpasses ViT and CNN with a 0.9811 AUC and 0.9858 accuracy score. Our network outperforms previous techniques to reduce intra-class variability and generate well-separated feature embedding. The CAM figure illustrates that the decision region corresponds to radiologists' detecting spots. The suggested method can be an effective way of catching COVID-19 instances.

Recent advances of Transformers in medical image analysis: A comprehensive review

“…LAT 87 adopted the encoder‐to‐decoder structure, in which a pixel relation based encoder and a lesion filter based decoder were set. The performance of LAT 137 was tested on Messidor‐1, Messidor‐2, and EyePACS. Figure 5 showed the structure of LAT in Wang et al 137 …”

“…The performance of LAT 137 was tested on Messidor‐1, Messidor‐2, and EyePACS. Figure 5 showed the structure of LAT in Wang et al 137 …”

“…Considering the serious damage of COVID-19 on global health and economy, it is urgent to achieve a fast and effective diagnosis of COVID-19. Thus, authors in Refs, 75,[90][91][92][93][94][95][96][122][123][124][125][127][128][129]133,135,137 explored to achieve a rapid and accurate COVID-19 classification using Transformerbased architectures. Shome et al 93 proposed a Covid-Transformer to achieve the automatic examining of COVID-19 according to the X-ray image.…”

Multisemantic Level Patch Merger Vision Transformer for Diagnosis of Pneumonia

“…In 2020, Dosovitskiy et al proposed the Vision Transformer (ViT) model [ 13 ], which achieves comparable or even better classification performance than CNN with lower computational cost, making ViT model a hot research topic in CV field. Some studies have tried to apply it on the medical image tasks, e.g., comparing the performance between ViT and some other models and proposing to use ViT for pneumonia diagnosis [ 14 ], or improving ViT for COVID-19 detection [ 15 ]. To reduce the computational efforts of ViT, in 2022, Renggli et al [ 16 ] proposed the Patch Merger, which significantly reduces the computational efforts of ViT while maintaining the model performance basically unchanged.…”

DataSheet_2.pdf

“…Participation in a clinical study of other drugs within 28 days before vaccination or planned participation within 6 months after vaccination;(8)Have an inherited bleeding tendency or coagulation disorder (eg: Cytokine deficiency, coagulopathy, or thrombocytopenia), or a history of severe bleeding;(9)…”