Skip to main content

Table 2 Univariate and multivariate logistic regression analyses for selecting clinical and conventional imaging features

From: MRI feature-based radiomics models to predict treatment outcome after stereotactic body radiotherapy for spinal metastases

 

Univariate analysis

 

Multivariate analysis

 

OR (95% CI)

p-value

OR (95% CI)

p-value

Clinical features

 Age (years)

1.008 (0.986–1.030)

0.483

  

 Gender (n, %)

0.720 (0.376–1.377)

0.320

  

 Primary histology (n, %)

0.925 (0.619–1.383)

0.704

  

 No. of spinal lesions (n, %)

2.335 (1.201–4.542)

0.012*

2.135 (1.010–4.514)

0.047*

 Lymphatic metastasis (n, %)

2.255 (1.110–4.580)

0.025*

1.870 (0.851–4.109)

0.119

 Visceral metastasis (n, %)

2.185 (0.720–6.631)

0.168

  

 Dose/fraction (n, %)

1.220 (0.829–1.796)

0.314

  

 Pain score

1.904 (1.164–3.113)

0.010*

1.970 (1.163–3.339)

0.012*

 KPS score (n, %)

2.509 (1.288–4.889)

0.007*

2.095 (1.019–4.307)

0.044*

 SINS

1.871 (1.039–3.369)

0.037*

1.401 (0.695–2.824)

0.346

Conventional imaging features

 Location (n, %)

1.278 (0.850–1.923)

0.239

  

 Bone destruction pattern (n, %)

1.226 (0.809–1.856)

0.337

  

 Soft-tissue mass (n, %)

1.334 (0.587–3.034)

0.491

  

 Vertebral body involvement (n, %)

0.909 (0.272–3.035)

0.877

  

 Pedicle involvement (n, %)

1.464 (0.756–2.834)

0.258

  

 Lamina involvement (n, %)

1.231 (0.645–2.351)

0.528

  

 Compression fracture (n, %)

2.069 (1.012–4.230)

0.046*

1.447 (0.653–3.207)

0.363

 Bilsky scale (n, %)

2.620 (1.338–5.128)

0.005*

2.481 (1.178–5.225)

0.017*

  1. KPS Karnofsky performance status, SINS Spine instability neoplastic score; *p < 0.05