Table 5.

Patch-wise comparisons of the accuracy, sensitivity, and specificity metrics for the ICIAR2018 dataset. The best results are shown in bold.

Class	Model	Accuracy		Sensitivity
				Without Data				Data
				Augmentation				Augmentation
				Non-Carcinoma		Carcinoma		Non-Carcinoma		Carcinoma
Binary	DNet	0.8233	0.8043	0.76		0.87		0.68		0.94
	MSI-MFNet	0.83	0.82	0.78		0.89		0.72		0.96
				Benign	Normal	In Situ	Invasive	Benign	Normal	In Situ	Invasive
Multi	DNet	0.601	0.631	0.300	0.83	0.763	0.710	0.264	0.927	0.805	0.75
	MSI-MFNet	0.64	0.68	0.31	0.717	0.90	0.83	0.39	0.932	0.82	0.82
				Specificity
				Without Data				Data
				Augmentation				Augmentation
				Non-Carcinoma		Carcinoma		Non-Carcinoma		Carcinoma
Binary	DNet			0.87		0.78		0.94		0.68
	MSI-MFNet			0.89		0.79		0.96		0.72
				Benign	Normal	In Situ	Invasive	Benign	Normal	In Situ	Invasive
Multi	DNet			0.879	0.809	0.94	0.819	0.963	0.788	0.892	0.853
	MSI-MFNet			0.95	0.88	0.872	0.83	0.97	0.86	0.90	0.86