3.
Logistic regression prediction models developed in development cohort
| Variables | RPRS | Clinical model | |||||
| Coefficient | OR (95% CI) | P | Coefficient | OR (95% CI) | P | ||
| CT, computed tomography; LN, lymph node; RPRS, radiomics-based predictive risk score; OR, odds ratio; 95% CI, 95% confidence interval; NA, not applicable. | |||||||
| Age group | −0.659 | 0.517
(0.316, 0.846) |
0.009 | −0.622 | 0.537
(0.333, 0.866) |
0.011 | |
| Tumor diameter | 0.581 | 1.788
(1.077, 2.969) |
0.025 | 0.750 | 2.117
(1.294, 3.463) |
0.003 | |
| Tumor location | 1.624 | 5.073
(2.835, 9.079) |
<0.001 | 1.742 | 5.710
(3.219, 10.128) |
<0.001 | |
| CT-based LN status | 0.678 | 1.970
(1.184, 3.276) |
0.009 | 0.923 | 2.517
(1.543, 4.107) |
<0.001 | |
| Radiomics-based classifier | 1.514 | 4.546
(2.347, 8.806) |
<0.001 | NA | NA | NA | |