Table 2 Prediction performance of different models.

From: A prognostic model integrating radiomics and deep learning based on CT for survival prediction in laryngeal squamous cell carcinoma

Model

C-index (95% CI)

1-year AUC (95% CI)

2-year AUC (95% CI)

3-year AUC (95% CI)

Training set

    

Clinical model

0.683(0.623–0.743)

0.732(0.636–0.829)

0.731(0.641–0.821)

0.722(0.639–0.805)

RS

0.711(0.648–0.775)

0.780(0.675–0.885)

0.773(0.691–0.855)

0.760(0.683–0.837)

DLS

0.742(0.675–0.809)

0.751(0.626–0.874)

0.756(0.658–0.854)

0.781(0.697–0.865)

Combined model

0.826(0.779–0.873)

0.852(0.765–0.940)

0.867(0.803–0.932)

0.888(0.835–0.941)

Internal testing set

    

Clinical model

0.634(0.528–0.740)

0.661(0.455–0.868)

0.686(0.529–0.843)

0.679(0.547–0.811)

RS

0.679(0.571–0.787)

0.638(0.432–0.844)

0.693(0.530–0.856)

0.742(0.618–0.867)

DLS

0.727(0.624-0.832)

0.787(0.595-0.979)

0.809(0.670-0.948)

0.748(0.614-0.883)

Combined model

0.810(0.736-0.883)

0.820(0.685-0.955)

0.880(0.792-0.969)

0.858(0.768-0.949)

External testing set

    

Clinical model

0.602(0.493-0.711)

0.589(0.330-0.848)

0.597(0.433-0.761)

0.622(0.487-0.756)

RS

0.617(0.509-0.725)

0.693(0.534-0.852)

0.670(0.508-0.831)

0.658(0.514-0.803)

DLS

0.729(0.623-0.835)

0.652(0.378-0.927)

0.784(0.623-0.944)

0.709(0.574-0.844)

Combined model

0.742(0.649-0.834)

0.707(0.521-0.893)

0.812(0.683-0.942)

0.778(0.651-0.904)

  1. RS, radiomics score; DLS, deep learning score; AUC, area under curve; CI, confidence interval.