Skip to main content
. 2023 Feb 27;59(3):469. doi: 10.3390/medicina59030469

Figure 4.

Figure 4

(A) The plot shows receiver operating characteristics (ROC) curve analyses with artificial intelligence-calculated hepatorenal index (AI-HRI) for the classification of normal liver and mild steatosis. The accuracy of AI-HRI was fair based on the area under the curve (AUC) value of 0.758. The 95% confidence intervals of the AUC are listed inside the brackets. (B) The plot shows that AI-HRI could detect moderate/severe steatosis with an AUC of 0.803, indicating good classification accuracy. The magnetic resonance imaging fat fraction (MRI-PDFF) was used as the reference method in both classifications.