Research Paper Volume 14, Issue 6 pp 2775—2792

Construction and evaluation of a nomogram for predicting survival in patients with lung cancer


Figure 6. Combination of the PRS signature and clinical features improves survival prediction in training sets. (A) A decision tree was constructed to improve risk stratification. (B) Multivariate Cox regression model (complex model). (C) Survival nomogram for quantifying risk assessment for individual patients. (D) Calibration analysis revealed a high degree of accuracy in predicting survival at 3 or 5 years. (E) Among all clinical variables, tROC analysis demonstrated that the nomogram was the most stable and powerful predictor of OS.