. 2022 Mar 19;30:100916. doi: 10.1016/j.imu.2022.100916

Table 5.

Various scores calculated in test dataset for different ResNet50 TL model where Pre $=$ Precision, Re $=$ Recall, F1 $=$ F1-score, Sup $=$ Support, Acc $=$ Accuracy, INCXR14 $= I m a g e N e t_C h e s t X - r a y 14$ , INCxP $= I m a g e N e t_C h e x P e r t$ , iNSup $= i N a t 2021_S u p e r v i s e d$ , iNSupFS $= i N a t 2021_S u p e r v i s e d_F r o m_S c r a t c h$ , iNMSwAV $= i N a t 2021_M i n i_S w A V_1 k$ .

Model	Class	Pre	Re	F1	Sup	Acc
ChestX-ray14	Covid	0.9791	0.9737	0.9764	723	0.9766
ChestX-ray14	Normal	0.9741	0.9795	0.9768	730	0.9766
ChexPert	Covid	0.9861	0.9834	0.9848	723	0.9849
ChexPert	Normal	0.9836	0.9863	0.9850	730	0.9849
ImageNet	Covid	0.8394	0.7953	0.8168	723	0.8224
ImageNet	Normal	0.8073	0.8493	0.8278	730	0.8224
INCXR14	Covid	0.9902	0.9737	0.9819	723	0.9821
INCXR14	Normal	0.9744	0.9904	0.9823	730	0.9821
INCxP	Covid	0.9650	0.9544	0.9597	723	0.9601
INCxP	Normal	0.9553	0.9658	0.9605	730	0.9601
iNSup	Covid	0.9635	0.9488	0.9561	723	0.9566
iNSup	Normal	0.9501	0.9644	0.9572	730	0.9566
iNSupFS	Covid	0.9696	0.9710	0.9703	723	0.9704
iNSupFS	Normal	0.9712	0.9699	0.9705	730	0.9704
iNMSwAV	Covid	0.9931	0.9903	0.9917	723	0.9917
iNMSwAV	Normal	0.9904	0.9932	0.9918	730	0.9917
$M o C o_v 1$	Covid	0.9411	0.9281	0.9345	723	0.9353
$M o C o_v 1$	Normal	0.9297	0.9425	0.9361	730	0.9353
$M o C o_v 2$	Covid	0.7974	0.8382	0.8173	723	0.8135
$M o C o_v 2$	Normal	0.8312	0.7891	0.8096	730	0.8135