Skip to main content

Table 3 Performance obtained by the proposed methods on the testing set

From: Breast DCE-MRI: lesion classification using dynamic and morphological features by means of a multiple classifier system

Classifier

Segmentation

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

Accuracy (%)

p valuea

Bayesian classifier using dynamic features

Manual

92.3 (24/26)

81.8 (18/22)

85.7 (24/28)

90.0 (18/20)

87.5

0.04

Automatic

88.5 (23/26)

68.2 (15/22)

76.7 (23/30)

83.3 (15/18)

79.2

Decision tree classifier using morphological features

Manual

92.3 (24/26)

77.3 (17/22)

82.8 (24/29)

89.5 (17/19)

85.4

0.02

Automatic

76.9 (20/26)

40.9 (9/22)

60.6 (20/33)

60.0 (9/15)

60.4

 
  1. PPV positive predictive value, NPV negative predictive value
  2. aMcNemar test