A CT-Based Radiomics Nomogram Model for Differentiating Primary Malignant Melanoma of the Esophagus from Esophageal Squamous Cell Carcinoma

Figure 3

Radiomics nomogram model based on radiomics signature for differentiating PMME from ESCC. (a) The developed radiomics nomogram. (b) Calibration curves of radiomics nomogram model in primary cohort. (c) Calibration curves of radiomics nomogram model in validation cohort. The axis showed the actual probability of PMME. The axis showed the predicted probability of PMME. The black dotted line (ideal) presented a perfect prediction by an ideal model. The red dotted line (apparent) showed calibration curve of the nomogram model in this study. The blue line (bias-corrected) showed estimation of predicted values by bootstrapping. Calibration curves presented the calibration of radiomics nomogram model in terms of the agreement between the predicted probability of PMME and the actual outcomes of the PMME. The (b) and (c) suggested a better prediction of radiomics nomogram model for diagnosing PMME in primary and validation cohorts.

(a)

(b)

(c)