Table 3.

Detection performance results for the proposed CNN models with a 75/25 training/testing ratio on the third dataset.

Models	Evaluation metric
	ACC	SEN	SPE	PRECI	F1_score	MCC	Error	Kappa	FPR
InceptionResNetv2	0.9446	0.9988	0.8904	0.9011	0.9474	0.8944	0.0554	0.8892	0.1096
Inceptionv3	0.9767	0.9855	0.9679	0.9685	0.9769	0.9535	0.0233	0.9534	0.0321
ResNet50	0.9691	0.9970	0.9412	0.9443	0.9699	0.9397	0.0309	0.9382	0.0588
BRAIN-TUMOR-net	1.00	1.00	1.00	1.00	1.00	1.00	0.0	1.00	0.0