Differentiation of the Follicular Neoplasm on the Gray-Scale US by Image Selection Subsampling along with the Marginal Outline Using Convolutional Neural Network

Table 7

Result of the CNN inference conducted on the test data groups, both and .

	Predicted_Adenoma	Predicted_Carcinoma		Overall accuracy
	True_Adenoma; 100%	False_Carcinoma; 0.00%	True negative rate; 1.0	100%
	False_Adenoma; -	True_Carcinoma; -	True positive rate; -
	False omission rate; -	Positive predictive value; -	-score: -
			-score: -
			-score: -
			-mean: -
	True_Adenoma; 69.76%	False_Carcinoma; 30.24%	True negative rate; 0.6976	70.37%
	False_Adenoma; 28.95%	True_Carcinoma; 71.05%	True positive rate; 0.7105
	False omission rate; 0.2683	Positive predictive value; 0.6749	-score: 0.6818
			-score: 0.6923
			-score: 0.7031
			-mean: 0.6925
Total	True_Adenoma; 93.19%	False_Carcinoma; 6.81%	True negative rate; 0.9319	89.52%
	False_Adenoma; 28.95%	True_Carcinoma; 71.05%	True positive rate; 0.7105
	False omission rate; 0.0582	Positive predictive value; 0.6750	-score: 0.6818
			-score: 0.6923
			-score: 0.7031
			-mean: 0.6925