Table 4.

The output metrics for the convolutional neural networks. Ensemble_A is comprised of a combination of all five of the convolutional neural networks. Ensemble_B consists of three models that produced the best outputs together, which in this case was the Inception V3, Resnet, and Xception with drop/aux. Output metrics were obtained using the validation-test data set

Model	Views	Accuracy	Sensitivity	Specificity	PPV	NPV
Inception V3	One	0.70	0.68	0.73	0.71	0.70
	Three	0.74	0.73	0.75	0.74	0.73
Resnet	One	0.73	0.68	0.77	0.75	0.71
	Three	0.75	0.70	0.80	0.78	0.73
Resnet with drop/aux	One	0.72	0.74	0.70	0.71	0.73
	Three	0.73	0.73	0.73	0.73	0.73
Xception	One	0.75	0.73	0.76	0.75	0.74
	Three	0.78	0.75	0.80	0.79	0.76
Xception with drop/aux	One	0.75	0.71	0.80	0.78	0.73
	Three	0.78	0.73	0.73	0.81	0.75
Ensemble_A	One	0.76	0.77	0.76	0.76	0.77
	Three	0.81	0.80	0.83	0.82	0.81
Ensemble_B	One	0.75	0.68	0.82	0.79	0.72
	Three	0.80	0.73	0.88	0.85	0.76