Table 2.

AI techniques applied in the diagnosis of COVID-19

AI field	Biomedical imaging	Techniques	Results
Deep learning	CT	Lesion-attention deep neural network [81]	88.6% ACC
			94% AUC
			88.8% RE
			87.9% PRE
		Multi-Task deep learning [80]	98.78% ACC
	X-ray	Deep transfer learning [88]	98% RE ± 3
			90% SP
		EfficientNet family of models [82]	93.9% ACC
			96.8% RE
		Faster R–CNN [83]	97.36% ACC
			97.65% RE
			99.28% PRE
		Inception V3 [16]	>96% ACC
		Multi-Task deep learning [80]	84.67% ACC
Machine learning	CT	Support Vector Machine [89]	99.68% ACC
	X-ray	Support Vector Machine [84]	100% ACC
		Support Vector Machine [90]	99.27% ACC