Skip to main content
Insights into Imaging

Table 2 Univariate and multivariable logistic regression analyses for selecting clinical features of model development

From: Feasibility and effectiveness of automatic deep learning network and radiomics models for differentiating tumor stroma ratio in pancreatic ductal adenocarcinoma

Characteristics

Univariate analysis

Multivariate analysis

OR (95% CI)

p-value

OR (95% CI)

p-value

Age

1.004 (0.972, 1.038)

0.793

  

Gender

1.112 (0.575, 2.15)

0.753

  

Abdominal pain

1.84 (0.944, 3.586)

0.074

  

Pancreatitis history

0.612 (0.285, 1.314)

0.208

  

Jaundice

0.822 (0.395, 1.708)

0.599

  

T stage

0.416 (0.192, 0.902)

0.026*

0.410 (0.205, 0.821)

0.012*

Histological grade

1.729 (0.895, 3.342)

0.103

  

Lymph node metastasis

1.102 (0.576, 2.111)

0.769

  

Duodenum Invasion

1.32 (0.658, 2.648)

0.435

  

CT-reported tumor size

0.978 (0.951, 1.006)

0.121

  

Location

0.687 (0.331, 1.429)

0.316

  

Parenchymal atrophy

0.835 (0.424, 1.646)

0.602

  

PD dilatation

1.044 (0.452, 2.41)

0.92

  

CBD dilatation

0.558 (0.227, 1.371)

0.203

  

CA-199 level

1.281 (0.65, 2.526)

0.474

  

CEA level

1.103 (0.467, 2.606)

0.824

  

TBIL level

0.558 (0.227, 1.371)

0.149

  
  1. OR odds ratio, CI confidence interval, PD pancreatic duct, CBD common bile duct, CA-199 carbohydrate antigen 199, CEA carcino-embryonic antigen, TBIL total bilirubin
  2. *Represents p < 0.05
Back to article page