Table 5.

- Classification Performance results for test and validation sets: Accuracy, precision, recall, F1-Score and ROC-AUC.

Metric $\mathit{\mu }\pm 2\mathit{\sigma }\%$	Four Classes		Seven Classes
	ResNet18	ResNet50	ResNet18	ResNet50
Training Accuracy	$96.70\pm 0.01$	$98.32\pm 0.02$	$96.76\pm 0.01$	$98.72\pm 0.01$
Training Precision	$96.27\pm 0.08$	$96.65\pm 0.20$	$96.82\pm 0.07$	$97.57\pm 0.12$
Training Recall	$96.09\pm 0.07$	$97.23\pm 0.15$	$96.17\pm 0.08$	$98.62\pm 0.05$
Training F1-Score	$96.19\pm 0.07$	$98.27\pm 0.04$	$95.43\pm 0.06$	$99.22\pm 0.02$
Training ROC-AUC	$99.70\pm 0.01$	$99.95\pm 0.01$	$99.76\pm 0.01$	$99.97\pm 0.01$
Test Accuracy	$97.64\pm 1.79$	$98.43\pm 1.38$	$99.33\pm 0.59$	$99.72\pm 0.26$
Test Precision	$97.47\pm 1.99$	$98.59\pm 1.36$	$99.50\pm 0.43$	$99.41\pm 0.53$
Test Recall	$97.36\pm 1.81$	$98.23\pm 1.44$	$98.51\pm 1.29$	$98.93\pm 0.98$
Test F1-Score	$97.37\pm 1.92$	$98.45\pm 1.51$	$98.45\pm 1.49$	$98.94\pm 0.81$
Test ROC-AUC	$97.72\pm 0.63$	$99.91\pm 0.07$	$99.94\pm 0.02$	$99.93\pm 0.03$
Test Accuracy	$97.64\pm 1.79$	$98.43\pm 1.38$	$99.33\pm 0.59$	$99.72\pm 0.26$
Validation Accuracy	$97.18\pm 1.40$	$97.93\pm 1.20$	$99.37\pm 0.60$	$97.73\pm 1.46$
Validation Precision	$96.70\pm 1.80$	$97.82\pm 1.60$	$98.52\pm 1.40$	$94.71\pm 3.20$
Validation Recal	$96.95\pm 1.61$	$97.52\pm 1.21$	$98.44\pm 1.41$	$94.16\pm 0.74$
Validation F1-Score	$96.76\pm 1.82$	$97.66\pm 1.41$	$98.13\pm 1.80$	$93.73\pm 4.41$
Validation ROC-AUC	$99.78\pm 0.20$	$99.81\pm 0.18$	$99.95\pm 0.03$	$99.78\pm 0.20$