Table 4. Diagnostic performance of multi-parametric models.

Disease groups	Accuracy	AUC	Sensitivity	Specificity
No vs. acute vs. chronic	0.65	0.79	0.77	0.69
No vs. chronic	0.76	0.82	0.86	0.61
No vs. F1	0.74	0.84	0.75	0.70

Predicting factors used for the multiparametric approach were: (I) patient age; (II) hepatic fat fraction; (III) presence of liver iron overload quantification; (IV) hepatobiliary phase parenchymal gadolinium [concentration] and (V) presence of ascites. AUC, area under the curve.