Skip to main content

Table 2 Evaluation measures for caries classification according to tooth area segmentation

From: Tooth caries classification with quantitative light-induced fluorescence (QLF) images using convolutional neural network for permanent teeth in vivo

Evaluation metrics

Segmentation for teeth area

No

Yes

True-positive case, n (%)

444 (52.5)

423 (52.0)

False-positive case, n (%)

75 (8.9)

53 (6.5)

True-negative case, n (%)

259 (30.7)

273 (33.6)

False-negative case, n (%)

67 (7.9)

64 (7.9)

Sensitivity

0.869

0.869

Specificity

0.775

0.810

Negative predictive value

0.795

0.837

f1-scores

0.862

0.879

Precision

0.856

0.889

Accuracy

0.832

0.856

AUC

0.909

0.928