Table 5 The discrimination performance of models for PLNM prediction in the held-out cohort

Model 1

0.89 (0.85, 0.94)

0.79 (0.73, 0.85)

79.07 (64.01, 90.02)

89.09 (83.34,93.41)

7.25 (4.62, 11.51)

0.23 (0.12,0.43)

Model 2

0.90 (0.85, 0.94)

0.81 (0.75, 0.86)

88.37 (74.91, 96.12)

83.64 (77.13, 88.94)

5.40 (3.81, 7.82)

0.14 (0.06, 0.30)

Model 3

0.71 (0.69, 0.78)

0.74 (0.68, 0.79)

69.77 (53.91, 82.8)

75.15 (67.80, 81.5)

2.81 (2.01, 3.92)

0.40 (0.32, 0.61)

Model 4

0.78 (0.77, 0.88)

0.84 (0.78, 0.88)

81.40 (66.63, 91.63)

84.24 (77.80, 89.41)

5.17 (3.54, 7.62)

0.22 (0.13, 0.42)

For LNs with short diameter ≤ 10 mm

Model 1

0.78 (0.70, 0.84)

0.82 (0.75, 0.88)

84.62 (54.61, 98.12)

70.45 (61.90, 78.12)

2.86 (2.01, 4.12)

0.22 (0.06, 0.82)

Model 2

0.83 (0.76, 0.89)

0.84 (0.77, 0.89)

76.92 (46.2, 95.00)

84.85 (77.6, 90.5)

5.08 (3.1, 8.4)

0.27 (0.1, 0.7)

Model 3

0.66 (0.52, 0.77)

0.65 (0.53, 0.76)

73.33 (54.12, 87.73)

57.78 (39.2, 74.5)

1.73 (1.12, 2.74)

0.46 (0.23, 0.91)

Model 4

0.74 (0.66, 0.88)

0.78 (0.66, 0.86)

86.67 (69.31, 96.25)

69.70 (51.32, 84.46)

2.86 (1.72, 4.93)

0.19 (0.07, 0.51)

For LNs with short diameter > 10 mm

Model 1

0.91 (0.82, 0.97)

0.73 (0.61, 0.82)

80.00 (61.42, 92.31)

90.91 (75.73, 98.12)

8.80 (2.91, 26.32)

0.22 (0.15, 0.52)

Model 2

0.92 (0.80, 0.96)

0.81 (0.69, 0.89)

83.33 (65.3, 94.4)

84.85 (68.1, 94.9)

5.50 (2.4, 12.5)

0.2 (0.09, 0.4)

Model 3

0.73 (0.62, 0.78)

0.78 (0.70, 0.84)

61.54 (31.6, 86.1)

79.55 (71.7, 86.1)

3.01 (1.71, 5.23)

0.48 (0.22, 1.00)

Model 4

0.79 (0.71, 0.85)

0.86 (0.79, 0.91)

69.23 (38.6, 90.9)

87.88 (81.1, 92.9)

5.71 (3.2, 10.3)

0.35 (0.21, 0.83)