Skip to main content

Table 1 Clinical characteristics across the training, internal validation, and independent test sets (test set 1 and test set 2) in patients

From: Predicting occult lymph node metastasis in solid-predominantly invasive lung adenocarcinoma across multiple centers using radiomics-deep learning fusion model

Characteristics

Training Set

Internal Validation Set

Independent Test Set 1

Independent Test Set 2

P-value

pN+ (n = 235)

pN- (n = 235)

pN+ (n = 101)

pN- (n = 101)

pN+ (n = 43)

pN- (n = 184)

pN+ (n = 99)

pN- (n = 327)

Agea (years)

59.9 ± 9.2

59.1 ± 9.9

60.0 ± 9.7

61.7 ± 9.0

59.4 ± 10.6

61.9 ± 9.6

59.7 ± 9.4

59.7 ± 9.7

0.048

Gender

        

< 0.001

 Male

108 (46.0)

101 (43.0)

47 (46.5)

42 (41.6)

21 (48.8)

90 (48.9)

46 (46.5)

146 (44.6)

 

 Female

127 (54.0)

134 (57.0)

54 (53.5)

59 (58.4)

22 (51.2)

94 (51.1)

53 (53.5)

181 (55.4)

 

Maximum Diametera (cm)

1.8 ± 0.4

1.6 ± 0.4

1.8 ± 0.4

1.6 ± 0.3

1.8 ± 0.4

1.8 ± 0.6

1.8 ± 0.5

1.6 ± 0.4

0.040

Density

        

< 0.001

 Solid

201 (85.5)

198 (84.3)

85 (84.2)

88 (87.1)

39 (90.7)

137 (74.5)

83 (83.8)

239 (73.1)

 

 Part-Solid

34 (14.5)

37 (15.7)

16 (15.8)

13 (12.9)

4 (9.3)

47 (25.5)

16 (16.2)

88 (26.9)

 

Location

        

< 0.001

 Inner

21 (8.9)

11 (4.7)

12 (11.9)

5 (5.0)

8 (18.6)

8 (4.3)

8 (8.1)

17 (5.2)

 

 Middle

59 (25.1)

54 (23.0)

29 (28.7)

34 (33.7)

9 (20.9)

41 (22.3)

24 (24.2)

106 (32.4)

 

 Outer

155 (66.0)

170 (72.3)

60 (59.4)

62 (61.4)

26 (60.5)

135 (73.4)

67 (67.7)

204 (62.4)

 
  1. Unless otherwise specified, the data presented were the number of patients, with the corresponding percentages denoted in parentheses. The p-values represented the outcomes of univariable association analysis conducted for each characteristic across four datasets
  2. SD Standard deviation
  3. aData were means ± SDs