Skip to main content
. 2017 Jan 6;8(8):13387–13399. doi: 10.18632/oncotarget.14534

Table 4. Logistic regression analysis of suspicious conventional US features and SWS imagine indices for predicting malignancy.

Parameter β SE OR (95% CI) P Value
Univariate analysis
Conventional US risk factor
 Poorly-defined Margin 2.342 0.293 10.407 (5.8585,18.486) 0.000*
 Microcalcification 1.599 0.255 4.950 (3.005, 8.154) 0.000*
 Solid Component 1.504 0.349 4.501 (2.270,8.924) 0.000*
 Taller-than-wide shape 1.186 0.262 3.272 (1.959,5.466) 0.000*
 Hypoechogenicity 0.976 0.253 2.654 (1.616,4.360) 0.000*
SWS imagine risk factor
 E-mean ≥ 2.46 m/s 2.108 0.283 8.229 (4.727,14.327) 0.000*
 E-max ≥ 3.52 m/s 1.074 0.137 2.296 (2.238,3.825) 0.000*
 E-mean ≥ 23.00 kpa 0.117 0.017 1.124 (1.088,1.161) 0.000*
 E-max ≥ 47.00 kpa 0.043 0.006 1.044 (1.032,1.056) 0.000*
Multivariate analysis
 Conventional US risk factor
 Poorly-defined Margin 2.053 0.439 7.792 (3.295,18.424) 0.000*
 Taller-than-wide shape 1.151 0.392 3.160 (1.466,6.812) 0.003*
 Microcalcification 0.885 0.367 2.422 (1.179, 4.975) 0.016*
SWS imagine risk factor
 E-max ≥ 3.52 m/s 2.819 1.010 16.760 (2.317,121.257) 0.005*
 E-max ≥ 47.00 kpa 0.090 0.046 0.914 (0.835,0.999) 0.045*

SWS imagine = shear wave speed imagine ; E-max = the max value of shear wave speed imagine; E-mean = the mean value of shear wave speed imagine

Note.- β, regression coefficient; SE, standard error; OR, odds ratio; CI, confidence interval