Assisting Scalable Diagnosis Automatically via CT Images in the Combat against COVID-19

Abstract

The pandemic of coronavirus Disease 2019 (COVID-19) caused enormous loss of life globally. 1-3 Case identification is critical. The reference method is using real-time reverse transcription PCR (rRT-PCR) assays, with limitations that may curb its prompt large-scale application. COVID-19 manifests with chest computed tomography (CT) abnormalities, some even before the onset of symptoms. We tested the hypothesis that application of deep learning (DL) to the 3D CT images could help identify COVID-19 infections. Using the data from 920 COVID-19 and 1,073 non-COVID-19 pneumonia patients, we developed a modified DenseNet-264 model, COVIDNet, to classify CT images to either class. When tested on an independent set of 233 COVID-19 and 289 non-COVID-19 patients. COVIDNet achieved an accuracy rate of 94.3% and an area under the curve (AUC) of 0.98. Application of DL to CT images may improve both the efficiency and capacity of case detection and long-term surveillance.

A Review on Deep Learning Techniques for the Diagnosis of Novel Coronavirus (COVID-19)

“…As far as performance is concerned, the system developed in [66] obtained 100% sensitivity. Among the reviewed systems, most of the frameworks [60] – [62] , [64] , [66] , [105] , [108] , [109] , and [111] achieved comparatively higher accuracy, sensitivity, specificity, precision, F1-score and AUC having these measure (where applicable) greater than 90%. The highest accuracy of 99.51% and 99.68% were found at [66] , [122] using pre-trained model and customized network respectively.…”

“…The experimental outcomes revealed that the scheme achieved accuracy, sensitivity, specificity, precision, and F1-score of 79.3%, 83%, 67%, 55%, and 63% respectively on the testing samples. Moreover, Liu et al [108] developed an automatic COVID-19 diagnosis system using deep learning method via CT images. The system used modified DenseNet-264 (COVIDNet) for diagnosis where the model consisted of 4 dense blocks.…”

Modality alignment contrastive learning for severity assessment of COVID-19 from lung ultrasound and clinical information

“…(2020) proposed a uncertainty vertex-weighted hypergraph learning method for identifying COVID-19 from CAP with image features and radiometric features. Other classification work includes detection of COVID-19 from negative cases ( Jin, Wang, Xu, Luo, Wei, Zhao, Hou, Ma, Xu, Zheng, et al., 2020b , Jin, Chen, Cao, Xu, Zhang, Deng, Zheng, Zhou, Shi, Feng, 2020a , Liu, Liu, Dai, Yang, Xie, Tan, Du, Shan, Zhao, Zhong, Lin, Guan, Xing, Sun, Wang, Zhang, Fu, Fan, Li, Zhang, Li, Liu, Xu, Du, Zhao, Hu, Fan, Wang, Wu, Nie, Cheng, Ma, Li, Jia, Liu, Guo, Huang, Shen, An, Li, Zhou, He, 2020a ), image-level and patient-level detection of COVID-19 ( Chen et al., 2020a ), radiomics models for prediction of hospital stay (long term or short term) ( Qi et al., 2020 ).…”

Deep Learning Fusion for COVID-19 Diagnosis

“…Considering CT imagery, Wang et al [32] modify the inception concept [20] to present a smaller feature dimension. Accordingly, Liu et al [33] alter DenseNet-264 [34] to present four dense blocks. Custom deep learning models also utilize X-ray imagery.…”

“…Accordingly, Liu et al . [33] alter DenseNet-264 [34] to present four dense blocks. Custom deep learning models also utilize X-ray imagery.…”