Framework for Classification of Chest X-Rays into Normal/COVID-19 Using Brownian-Mayfly-Algorithm Selected Hybrid Features

Table 3

Initial results achieved with pretrained deep learning methods.

Method	Pooling	TP	FN	TN	FP	AC	PR	SE	SP	F1S	NPV
AlexNet	MP	221	17	228	14	0.9354	0.9404	0.9286	0.9421	0.9345	0.9306
	AP	223	12	230	15	0.9437	0.9370	0.9489	0.9388	0.9429	0.9504
VGG16	MP	223	16	231	10	0.9458	0.9571	0.9331	0.9585	0.9449	0.9352
	AP	224	14	233	9	0.9521	0.9614	0.9412	0.9628	0.9512	0.9433
VGG19	MP	222	12	227	19	0.9354	0.9212	0.9487	0.9228	0.9347	0.9498
	AP	228	19	219	14	0.9313	0.9421	0.9231	0.9399	0.9325	0.9202
ResNet18	MP	230	16	218	16	0.9333	0.9350	0.9350	0.9316	0.9350	0.9316
	AP	231	13	222	14	0.9437	0.9429	0.9467	0.9407	0.9448	0.9447
ResNet50	MP	216	19	228	17	0.9250	0.9270	0.9191	0.9306	0.9231	0.9231
	AP	227	16	219	18	0.9292	0.9265	0.9342	0.9241	0.9303	0.9319
ResNet101	MP	229	13	220	18	0.9354	0.9271	0.9463	0.9244	0.9366	0.9442
	AP	219	15	230	16	0.9354	0.9319	0.9359	0.9350	0.9339	0.9388