Table 1.

Accuracy (%) of tested models for classifying COVID-19 versus non-COVID-19 CT images with different finding labels.

CNNs	Testing
	Baseline	$+\mathrm{GGO}$	$+\text{Consolidation}$	$+\mathrm{PE}$	$+\text{Combined}$
AlexNet	70.00	73.40*	75.90**	73.40*	78.80**
GoogleNet	72.40	75.90*	72.40	72.40	74.40*
DenseNet	80.80	79.30	79.30	80.80	77.80
NASNet-Mobile	80.30	82.30	78.80	80.80	82.30*
DarkNet	82.30	80.80	82.80	80.30	82.30
ResNet18	79.00	78.30	79.80	77.80	80.80

^*significant results with $p<0.05$

^**corrected $p$-value following Holm–Bonferroni

Note: Bold values represent the maximum value for each of CNN models.