Table 3 Data was divided in training/validation and test groups. During training/validation phase, three iterations were conducted, each with unique frame distribution.

From: Automated detection and classification of cervical and anal squamous cancer precursors using deep learning and multidevice colposcopy

 

Sensitivity (%)

Specificity (%)

PPV (%)

PPN (%)

Accuracy (%)

Fold 1

97.90

98.00

97.80

98.00

97.90

Fold 2

98.30

96.70

96.70

98.40

97.50

Fold 3

98.00

97.40

97.20

98.10

97.70

Average

98.10

97.40

97.20

98.10

97.70

IC95%

97.6–98.5

96.0–98.8

95.8–98.7

97.8–98.6

97.2–98.2

  1. In each iteration, the model was trained using two folds, and validate using the other one. Metrics were calculated in validation fold of each of iteration.
Back to article page