Skip to main content

Table 1 Demographic comparison between training and independent test cohorts

From: An integrated model incorporating deep learning, hand-crafted radiomics and clinical and US features to diagnose central lymph node metastasis in patients with papillary thyroid cancer

Characteristics

and US features

Training cohort

(n = 460)

Independent test cohort

(n = 153)

p value

Size (mean ± SD)

1.24 ± 0.92

1.16 ± 0.69

0.433

Sex

  

0.971

Male

136(75.1%)

45(24.9%)

 

Female

324(75.0%)

108(25.0%)

 

Age

  

0.484

≤ 55

415(75.5%)

135(24.5%)

 

≤55

45(71.4%)

18(28.6%)

 

Tumour location

  

0.711

Right lobe

228(73.8%)

81(26.2%)

 

Left lobe

213(76.6%)

65(23.4%)

 

Isthmus

19(73.1%)

7(26.9%)

 

Tumour position

  

0.095

Upper

152(81.7%)

34(18.3%)

 

Mid

164(71.9%)

64(28.1%)

 

Lower

125(72.3%)

48(27.7%)

 

Isthmus

19(73.1%)

7(26.9%)

 

Solid composition with hypoechoic echo

  

0.110

Present

417(74.2%)

145(25.8%)

 

Absent

43(84.3%)

8(15.7%)

 

Tumour multifocality

  

0.769

Present

83(76.1%)

26(23.9%)

 

Absent

377(74.8%)

127(25.2%)

 

Aspect ratio

  

0.198

Present

201(726%)

76(27.4%)

 

Absent

259(77.1%)

77(22.9%)

 

Microcalcification

  

0.769

Present

360(75.3%)

118(24.7%)

 

Absent

100(74.1%)

35(25.9%)

 

Tumour vascularity

  

0.320

0–1

401(74.4%)

138(25.6%)

 

2–3

59(79.7%)

15(20.3%)

 

Acoustic halo

  

0.804

Present

42(73.7%)

15(26.3%)

 

Absent

418(75.2%)

138(24.8%)

 

ETE

  

0.947

Present

119(74.8%)

40(25.2%)

 

Absent

341(75.1%)

113(24.9%)

 

Posterior acoustic attenuation

  

0.282

Present

67(79.8%)

17(20.2%)

 

Absent

393(74.3%)

136(25.7%)

 

CLNM in the pathology outcomes

  

0.982

Present

228 (75.0%)

76 (25.0%)

 

Absent

232 (75.1%)

77 (24.9%)

 
  1. Abbreviations: US, ultrasound; ETE, extrathyroidal extension; CLNM, central lymph node metastasis