Artificial Intelligence and Deep Learning Assisted Rapid Diagnosis of COVID-19 from Chest Radiographical Images: A Survey

Table 1

Performance analysis of several deep-learning-based models for diagnosis of COVID-19 from CT scan images.

Researchers	Deep learning model	Performance
Wang et al. [21]	Novel method (based on 3D U-Net++ [22] and ResNet50 [23])	0.99 AUC and 0.98 sensitivity
Zheng et al. [24]	DeCoVNet (based on U-Net [20])	90.1% accuracy
Song et al. [25]	DeepPneumonia	86% accuracy
Zhang et al. [26]	Novel method	92.49% accuracy
Li et al. [28]	COVNet (based on ResNet50 [23])	96% specificity and AUC of 0.96
Ardakani et al. [29]	Ten CNN models	99.51% accuracy using ResNet-101
Xu et al. [31]	Automated deep learning model (based on ResNet18)	86.7% accuracy
Chen et al. [32]	Deep learning model (based on U-Net++ [22])	98.85% accuracy
Shan et al. [33]	VB-Net (based on V-Net [34])	91.6% accuracy
Huang et al. [35]	Deep learning model (based on U-Net [20])	Quantification of CT parameters and analysis of lung opacities
Wang et al. [36]	Pretrained model	89.5% accuracy
Pathak et al. [37]	ResNet50 [23]	93% accuracy