Skip to main content

Table 2 Ultrasound and pathological features of DTC suspicious LNs

From: Dual-modal radiomics ultrasound model to diagnose cervical lymph node metastases of differentiated thyroid carcinoma: a two-center study

Variables

Training set

Internal test set

p1

External validation

p2

LN number

427

184

 

115

 

LN location

  

0.459

 

<0.001

 Middle neck

9 (2.1%)

7 (3.8%)

 

52 (45.2%)

 

 Left neck

204 (47.8%)

89 (48.4%)

 

6 (5.2%)

 

 Right neck

214 (50.1%)

88 (47.8%)

 

57 (49.6%)

 

LN regions

  

0.303

 

0.091

 I + II

18 (4.2%)

5 (2.7%)

 

6 (5.2%)

 

 III + IV + V

260 (60.9%)

104 (56.5%)

 

57 (49.6%)

 

 VI + VII

149 (34.9%)

75 (40.8%)

 

52 (45.2%)

 

Size(mm)

    

0.008

 Long axis

10 (7, 13)

9 (7, 12)

0.093

12 (8, 15)

 

 Short axis

5 (4, 6)

4 (3.75, 6)

0.474

6 (5, 7.5)

 

Oval shape

  

0.192

 

0.004

 Yes

173 (40.5%)

85 (46.2%)

 

64 (55.7%)

 

 No

254 (59.5%)

99 (53.8%)

 

51 (44.3%)

 

Hilum absence

  

0.971

 

<0.001

 Yes

300 (70.3%)

129 (70.1%)

 

103 (89.6%)

 

 No

127 (29.7%)

55 (29.9%)

 

12 (10.4%)

 

Focal hyperecho

  

0.235

 

0.001

 Yes

122 (28.6%)

44 (23.9%)

 

51 (44.3%)

 

 No

305 (71.4%)

140 (76.1%)

 

64 (55.7%)

 

Cystic degeneration

  

0.346

 

0.002

 Yes

30 (7%)

17 (9.2%)

 

19 (16.5%)

 

 No

397 (93%)

167 (90.8%)

 

96(83.5%)

 

Microcalcification

  

0.276

 

0.025

 Yes

66 (15.5%)

35 (19%)

 

28 (24.3%)

 

 No

361 (84.5%)

149 (81%)

 

87 (75.7%)

 

CDFI

  

0.574

 

0.002

 None

148 (34.7%)

57 (31%)

 

23 (20%)

 

 Hilar

80 (18.7%)

43 (23.4%)

 

21 (18.3%)

 

 Peripheral

70 (16.4%)

31 (16.8%)

 

16 (13.9%)

 

 Mixed

129 (30.2%)

53 (28.8%)

 

55 (47.8%)

 

Pathological diagnosis

  

0.967

 

<0.001

 Malignant

205 (48%)

88 (47.8%)

 

76 (66.1%)

 

 Benign

222 (52%)

96 (52.2%)

 

39 (33.9%)

 
  1. DTC: differentiated thyroid carcinoma; CDFI: color doppler flow imaging;
  2. p1: significance between training set and internal test set;p2 : significance between training set and external validation set