Figure 5
From: A Deep Learning-Based Radiomics Model for Prediction of Survival in Glioblastoma Multiforme
The nomogram (a) and calibration (b) curves. Radiomics signature and clinical data are associated with survival probability of 1, 2 and 3 years. The predictors are radiomics signature score, age of the patient (in years) and Karnofsky performance score (KPS). Draw a vertical line from each predictor to ‘Points’ to get the score of the predictor. Then summing up the scores of each predictor, the ‘Total Points’ correspond to the survival probability of 1, 2 and 3 years by drawing a vertical line from ‘Total Points’ to each survival probability axis. Calibration curves is used to assess the consistency between nomogram-predicted survival probability and actual fraction survival probability.
