Skip to main content

Table 5 Predictive performance of the clinical (age and menopause), radiomics, radiological, and clinical–radiomics models for delineating early stage from advanced-stage cervical cancer

From: Reduced field-of-view DWI‑derived clinical–radiomics model for the prediction of stage in cervical cancer

 

AUC

Sensitivity

Specificity

p value

Training set

Clinical model

0.673 (0.540, 0.787)

0.849 (0.681, 0.949)

0.429 (0.245, 0.628)

0.004

Radiomics model

0.867 (0.755, 0.940)

0.758 (0.577, 0.889)

0.893 (0.718, 0.977)

0.551

Radiological model

0.813 (0.692, 0.901)

0.697 (0.513, 0.844)

0.929 (0.765, 0.991)

0.206

Clinical–radiomics model

0.877 (0.767, 0.947)

0.697 (0.513, 0.844)

0.929 (0.765, 0.991)

Ref

Testing set

Clinical model

0.793 (0.617, 0.914)

0.790 (0.544, 0.939)

0.786 (0.492, 0.953)

0.095

Radiomics model

0.846 (0.678, 0.947)

0.790 (0.544, 0.939)

0.929 (0.661, 0.998)

0.433

Radiological model

0.823 (0.651, 0.933)

0.790 (0.544, 0.939)

0.857 (0.572, 0.982)

0.400

Clinical–radiomics model

0.887 (0.729, 0.970)

0.842 (0.604, 0.966)

0.786 (0.492, 0.953)

Ref

  1. Parentheses indicate 95% confidence interval
  2. AUC Area under the curve, Ref Reference
  3. p value means the difference of the models compared with the Ref based on DeLong test