Skip to main content

Table 5 Performance of the liver trauma segmentation model on the validation set stratified by AAST grade

From: Automated quantitative assessment of pediatric blunt hepatic trauma by deep learning-based CT volumetry

AAST grade

Dice (%)

Precision (%)

Recall (%)

RVD (%)

I (n = 3)

52.60 ± 24.02

60.90 ± 22.66

51.05 ± 34.74

-16.88 ± 47.58

II (n = 20)

67.06 ± 18.89

65.70 ± 21.79

72.74 ± 22.55

17.48 ± 42.74

III (n = 19)

78.42 ± 12.63

77.48 ± 18.75

82.47 ± 9.75

15.21 ± 42.70

IV (n = 8)

78.51 ± 6.93

77.48 ± 11.28

80.43 ± 8.16

5.630 ± 22.41

V (n = 2)

87.08 ± 4.33

86.59 ± 3.08

87.59 ± 5.60

1.111 ± 2.872

Overall (n = 52)

72.91 ± 16.75

72.40 ± 19.65

76.80 ± 18.55

12.22 ± 39.47