Skip to main content

Table 1 Clinical factors of the training and validation sets

From: An MR-based radiomics model for differentiation between hepatocellular carcinoma and focal nodular hyperplasia in non-cirrhotic liver

Clinical factors

Training set (n = 156)

P

Validation set (n = 68)

  

HCC (n = 104)

FNH (n = 52)

HCC (n = 45)

FNH (n = 23)

P

P*

Gender, male/female

84/20

18/34

0.001

40/5

12/11

0.002

0.136

Age, ≤50/>50 years

33/71

45/7

0.001

12/33

19/4

0.001

0.644

Hbsag, negative/positive

21/83

38/14

0.001

16/29

16/7

0.016

0.252

AFP, ≤400/> 400 ng/ml

99/5

52/0

0.261

45/0

23/0

-

0.317

MRI tumour number, single/multiple

89/15

36/16

0.028

40/5

16/7

0.101

0.838

MRI tumour size, cm

4.35 [2.60, 6.25]

2.50 [1.90, 3.33]

0.001

4.40 [3.00, 6.20]

2.70 [2.05, 3.95]

0.006

0.849

Liver steatosis, absent/present

78/26

39/13

1.000

33/12

20/3

0.331

0.762

Liver haemangioma, absent/present

88/16

49/3

0.141

42/3

22/1

1.000

0.235

Location, internal/subcapsular

36/68

30/22

0.010

16/29

11/12

0.474

0.829

Margin, ill-defined/well-defined

26/78

11/41

0.739

11/34

3/20

0.434

0.734

Shape, not round/round

46/58

29/23

0.234

18/27

16/7

0.040

0.905

Pseudocapsule, absent/present

67/37

28/24

0.270

24/21

13/10

1.000

0.448

Lesion homogeneity,

Heterogeneous/homogeneous

68/36

21/31

0.005

29/16

7/16

0.016

0.672

Lesion with steatosis, absent/present

86/18

49/3

0.082

38/7

22/1

0.337

0.895

Central vascular supply, absent/present

90/14

49/3

0.238

36/9

21/2

0.396

0.380

Central scar, absent/present

82/22

37/15

0.387

40/5

14/9

0.017

0.734

Enhancement pattern

  

0.001

  

0.001

0.56

 Early enhancement + washout

82

15

-

36

7

-

-

 Early enhancement + no washout

14

35

-

7

16

-

-

 Other patterns

8

2

-

2

0

-

-

  1. Note: HCC hepatocellular carcinoma in noncirrhotic liver, FNH focal nodular hyperplasia, HbsAg hepatitis B surface antigen, AFP alpha fetoprotein; P* represents the P value of comparison between training and validation set