Skip to main content

Table 5 Performance of logistic regression models on the five sub-datasets

From: Association of lower extremity peripheral arterial disease with quantitative muscle features from computed tomography angiography

Sub-dataseta

AUC (95% CI)

p value

Sensitivity

Specificity

Accuracy

Cut-point

Sub-dataset 1

 LRM-I

0.79 (0.68, 0.91)

< 0.001

60 (12/20)

75 (27/36)

70 (39/56)

0.35

 LRM-II

0.86 (0.76, 0.96)

< 0.001

70 (14/20)

81 (29/36)

77 (43/56)

0.43

Sub-dataset 2

 LRM-I

0.79 (0.66, 0.91)

< 0.001

70 (14/20)

64 (23/36)

66 (37/56)

0.35

 LRM-II

0.86 (0.76, 0.96)

< 0.001

75 (15/20)

83 (30/36)

80 (45/56)

0.43

Sub-dataset 3

 LRM-I

0.82 (0.70, 0.93)

< 0.001

80 (16/20)

78 (28/36)

79 (44/56)

0.35

 LRM-II

0.88 (0.79, 0.97)

< 0.001

75 (15/20)

81 (29/36)

79 (44/56)

0.43

Sub-dataset 4

 LRM-I

0.83 (0.72, 0.94)

< 0.001

85 (17/20)

69 (25/36)

75 (42/56)

0.35

 LRM-II

0.90 (0.83, 0.98)

 < 0.001

70 (14/20)

81 (29/36)

77 (43/56)

0.43

Sub-dataset 5

 LRM-I

0.78 (0.67, 0.90)

< 0.001

70 (14/20)

72 (26/36)

71 (40/56)

0.35

 LRM-II

0.86 (0.77, 0.96)

< 0.001

60 (12/20)

86 (31/36)

77 (43/56)

0.43

  1. Abbreviations: AUC Area under curve, CI Confidence interval, LRM Logistic regression model
  2. aThe segmentation region (from the inferior border of the patella to the superior border of the talus) of the lower leg muscles was divided into five equal segments. The CT images in each segment constituted an independent dataset, and sub-datasets 1–5 were constructed from the knee to the ankle