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

Table 2

Performance analysis of several deep learning models for classification of COVID-19 from X-ray images.

Researchers	Deep learning model	Performance
Zhang et al. [45]	Novel	96% accuracy
Panwar et al. [46]	nCOVnet	97.62% accuracy
Stephen et al. [48]	Novel	95% accuracy
Wang and Wong [49]	Novel	83.5% accuracy
Ghosal and Tucker [57]	ResNet50V2
Toğaçar et al. [54]	MobileNetV2 and SqueezeNet	99.27% accuracy
Ozturk et al. [56]	DarkCovidNet (based on YOLO)	98.08% accuracy
Narin et al. [58]	ResNet50 [23], InceptionV3, and InceptionResNetV2	98% accuracy using ResNet50
Hemdan et al. [62]	VGG19, DenseNet201, MobileNetV2, Xception, ResNetV2, InceptionV3, and InceptionResNetV2	90% accuracy using VGG19
Sethy et al. [61]	Thirteen CNN models	95.38% accuracy using RestNet50
Khan et al. [64]	CoroNet (based on Xception)	89.6% accuracy
Rahimzadeh and Attar [65]	The concatenated network of Xception and ResNet50V2	99.50% accuracy
Chowdhury et al. [66]	DenseNet201, ResNet18, AlexNet, and SqueezeNet	98.3% accuracy using SqueezeNet
Apostolopoulos et al. [68]	VGG19, MobileNetV2, InceptionResNetV2, Inception, and Xception	98.75% accuracy using VGG19
Altan and Karasu [69]	EfficientNetB0	99.69% accuracy
Brunese et al. [70]	Transfer learning	97% accuracy
Ucar and Korkmaz [71]	Deep-SqueezeNet (based on SqueezeNet)	98.26% accuracy
Mahmud et al. [72]	CovXNet	97.4% accuracy