Fig. 7From: Validation of multiparametric MRI based prediction model in identification of pseudoprogression in glioblastomasROC curve from a combination of imaging parameters (FA, CL, and rCBVmax) provided an AUC of 0.757 in identifying PsP and TP+mixed tumors. Leave-one-out cross-validation test revealed that 73.2% of cases were correctly classified as PsP and TP+mixed tumorsBack to article page