Establishment and Validation of a Gene Signature-Based Prognostic Model to Improve Survival Prediction in Adrenocortical Carcinoma Patients

Figure 4

Construction of an integrated model to optimize risk stratification and personalize risk assessment: (a) a decision tree was generated based on the TCGA cohort, with a risk score and SM status incorporated as two major components; (b) overall survival differed significantly in different risk subgroups, and mortality increased gradually as risk level increased; (c) a nomogram was constructed to personalize risk assessment for individual patients; (d) the calibration curve showed the nomogram-predicted survival to be close to ideal performance; and (e) tROC analysis of all clinicopathological parameters demonstrated the greatest accuracy of the nomogram for survival prediction.

(a)

(b)

(c)

(d)

(e)