Table 3.

Early Detection and Diagnosis publications (n=40) subclassified into 4 categories based on the modality used for COVID-19 prediction: (1) X-ray, (2) computed tomography, (3) lung ultrasound, and 4) nonimaging clinical data.

Category	Description	References	Publication count, n (%)
X-ray	Publications focused on the application of artificial intelligence techniques to detect and differentiate patients with COVID-19 from the general population using X-ray images.	[91, 92, 95, 97-99, 101-106, 112-114, 117, 118, 122-124, 126, 127, 129]	23 (57.5)
Computed tomography	Publications focused on the application of artificial intelligence techniques to detect and differentiate patients with COVID-19 from the general population using computed tomography images.	[93, 94,96, 100, 107-111, 116,119-121, 125, 130]	15 (37.5)
Lung ultrasound	Publication focused on the application of artificial intelligence techniques to detect and differentiate patients with COVID-19 from the general population using lung ultrasound images.	[128]	1 (2.5)
Nonimaging clinical data	Publication focused on the application of artificial intelligence techniques to detect and differentiate patients with COVID-19 from the general population using nonimaging clinical data.	[115]	1 (2.5)