Table 2.

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

Models	Evaluation metrics
	ACC	SEN	SPE	PRECI	F1_score	MCC	Error	Kappa	FPR
InceptionResNetv2	0.9689	0.9578	0.9800	0.9795	0.9685	0.9380	0.0311	0.9378	0.0200
Inceptionv3	0.9633	0.9622	0.9644	0.9644	0.9633	0.9267	0.0367	0.9267	0.0356
ResNet50	0.9656	0.9822	0.9489	0.9505	0.9661	0.9316	0.0344	0.9311	0.0511
BRAIN-TUMOR-net	0.9700	0.9644	0.9756	0.9753	0.9698	0.9401	0.0300	0.9400	0.0244