Table 3.

Testing accuracy for classifying mass and nonmass regions in INbreast dataset, where $\mu$ and $\sigma$ refer to the mean and standard deviation, respectively, for fivefold cross validation.

Model	Pretrained weight	Fine-tuning on INbreast	Testing accuracy ($\mu \pm \sigma$)	Training cascade
InceptionV3	ImageNet	Yes	$85.29\%\pm 4.29$	$\text{ImageNet}\to \text{INbreast}$
InceptionV3	CBIS-DDSM	Yes	$88.86\%\pm 2.96$	$\text{ImageNet}\to \mathrm{CBIS}\text{-}\mathrm{DDSM}\to \text{INbreast}$