Identification of Asymptomatic COVID-19 Patients on Chest CT Images Using Transformer-Based or Convolutional Neural Network-Based Deep Learning Models

Abstract

Novel coronavirus disease 2019 (COVID-19) has rapidly spread throughout the world; however, it is difficult for clinicians to make early diagnoses. This study is to evaluate the feasibility of using deep learning (DL) models to identify asymptomatic COVID-19 patients based on chest CT images. In this retrospective study, six DL models (Xception, NASNet, ResNet, EfficientNet, ViT, and Swin), based on convolutional neural networks (CNNs) or transformer architectures, were trained to identify asymptomatic patients with COVID-19 on chest CT images. Data from Yangzhou were randomly split into a training set (n = 2140) and an internal-validation set (n = 360). Data from Suzhou was the external-test set (n = 200). Model performance was assessed by the metrics accuracy, recall, and specificity and was compared with the assessments of two radiologists. A total of 2700 chest CT images were collected in this study. In the validation dataset, the Swin model achieved the highest accuracy of 0.994, followed by the EfficientNet model (0.954). The recall and the precision of the Swin model were 0.989 and 1.000, respectively. In the test dataset, the Swin model was still the best and achieved the highest accuracy (0.980). All the DL models performed remarkably better than the two experts. Last, the time on the test set diagnosis spent by two experts-42 min, 17 s (junior); and 29 min, 43 s (senior)-was significantly higher than those of the DL models (all below 2 min). This study evaluated the feasibility of multiple DL models in distinguishing asymptomatic patients with COVID-19 from healthy subjects on chest CT images. It found that a transformer-based model, the Swin model, performed best.

Keywords: Asymptomatic coronavirus-disease-2019 patients; Chest CT images; Convolutional neural networks; Deep learning; Transfer learning; Transformer.

Fig. 1

A flowchart of this study

Fig. 2

The performance of six models and two experts. A The six models’ performance in the validation dataset; B the times spent on the test dataset; C the six models’ and two experts’ performance in the test dataset

Fig. 3

A confusion matrix of two experts and six models. True positives, TP; true negatives, TN; false positives, FP; false negatives, FN

Fig. 4

ROC curves of six models with AUCs. ROC, receiver operating characteristic; AUC, area under ROC

Fig. 5

Two misclassified cases predicted by the Swin model. Swin, shifted window transformer. A Misdiagnosed as normal image with a probability of 0.682; B misdiagnosed as COVID-19 image with a probability of 0.505