Figure 14.6.
ROC chart comparing four risk models. Curves show true positive fraction versus false positive fraction for Briganti, Calibrated, Briganti plus, and Briganti double plus, with AUC values from zero point six nine to zero point seven nine.
Receiver Operating Curves (ROC) showing the discriminating performance of four risk models. The first risk model (Briganti) is the nomogram published online and recommended in the EAU guidelines. The second risk model is a calibrated logistic regression model (Calibrated). The third risk model (Briganti+) is a Random forest classifier that was trained using the same set of input variables as the Briganti risk model by means of a leave-on-patient-out training and testing scheme. The fourth risk model (Briganti++) is a similar Random forest classifier using additional input information from MRI (PI-RADS scores). The Area Under the Curve (AUC) was calculated for each risk model.

or Create an Account

Close Modal
Close Modal