Table 3 Results of the receiver operating characteristic analysis for the differentiation between benign (n = 86) and malignant (n = 105) breast lesions by ROI-analysis of indices IDf, ID, and If, whereby the analysed ROI-values indicate the amount of “malignant” voxels in the ROI.

From: Combined diffusion and perfusion index maps from simplified intravoxel incoherent motion imaging enable visual assessment of breast lesions

Test var

Malign

Benign

Asym 95% CI

AUC

Cut-off point

Sen

95% CI

Spec

95% CI

Acc

95% CI

MV ± SD

MV ± SD

IDf

86 ± 28

8 ± 22

0.899–0.973

0.936

38.8 *

0.924

0.872–0.976

0.930

0.880–0.980

0.927

0.876–0.978

ID

91 ± 20

25 ± 41

0.800–0.915

0.857

37.5 *

0.971

0.938–1.000

0.744

0.658–0.830

0.869

0.803–0.935

If

92 ± 20

29 ± 34

0.860–0.951

0.906

73.2 *

0.914

0.859–0.969

0.837

0.765–0.909

0.880

0.816–0.944

D′&f′

    

1,229.2/40.5

0.943

0.898–0.988

0.930

0.880–0.980

0.937

0.889–0.985

  1. The optimal cut-off point according the Youden index is given in 10–6 mm2/s for D′, in 10–3 for f′, and as percentages for IDf, ID, and If. Asymptotic significance of < 0.001 was reached for all parameter and index maps. Acc Accuracy, AUC Area under the curve, Asym 95% CI Asymptotic 95% confidence interval, Malign Malignant, MV Mean value, SD Standard deviation, Sens Sensitivity, Spec Specificity, Test var Test variable, * A higher test result indicates a more positive test.
  2. In addition, sensitivity, specificity, and accuracy are given for each index and for D′ and f′ in combination (D′&f′)7.
Back to article page