Skip to main content

Table 2 The features involved in prognostic models of OS, RFS and LR-RFS

From: CT imaging features associated with recurrence in non-small cell lung cancer patients after stereotactic body radiotherapy

  Features Level p-value Hazard Ratio validation
Point 95% CI Logistic regression 5-fold cross validation
Lower Upper AUC (95% CI) AUC (95% CI)
OS ECOG (ref = 0 or 1) 2 or 3 0.02 3.13 1.17 8.41 0.88 (0.78–0.97) 0.81 (0.80–0.82)
Vascular involvement (ref = 0) 1 0.01 3.21 1.29 8.03
Lymphadenopathy (ref = 0) 1 0.00 3.59 1.58 8.16
PC1a   0.04 1.24 1.02 1.51
RFS Vascular involvement (ref = 0) 1 0.01 3.06 1.40 6.70 0.86 (0.76–0.96) 0.80 (0.792–0.81)
Vascular attachment (ref = 0) 1 0.00 3.46 1.65 7.25
Pleural retraction (ref = 0) 1 0.01 3.24 1.41 7.42
Lymphadenopathy (ref = 0) 1 <.00 6.41 2.58 15.90
Relative enhancementa   0.05 1.40 1.00 1.96
LR-RFS F13 (9b_3D_Circularity)a   0.02 1.60 1.10 2.32 0.85 (0.74–0.95) 0.80 (0.78–0.81)
F92 (3D Laws features L5 L5 S5 Layer 1)a   0.00 1.96 1.35 2.83
Vascular involvement (ref = 0) 1 <.00 4.96 2.23 11.03
Lymphadenopathy (ref = 0) 1 0.02 2.64 1.19 5.82
  1. a: per 1standard deviation (SD) increase
  2. ECOG, Eastern cooperative oncology group; PC1: the 1st principle component