Accuracy of attenuation coefficient measurement (ACM) for hepatic steatosis: comparison with MRI proton density fat fraction (MRI-PDFF) and chemical fat analysis using multimodal liver fat phantoms.

Background: The objective of this study was to evaluate the accuracy and reproducibility of real-time ultrasound steatometry using attenuation coefficient measurement (ACM) in comparison with magnetic resonance imaging proton density fat fraction (MRI-PDFF) and chemical fat analysis. Specifically, we assessed the diagnostic performance and reproducibility of ultrasound-based ACM against MRI-PDFF and laboratory fat quantification using multimodal liver fat phantoms (LFPs) with varying fat-to-water ratios. Methods: Sixty LFPs with different fat concentrations were examined in two radiology centers and one chemical laboratory. Each phantom underwent three assessments: MRI-PDFF, ultrasound ACM, and laboratory-based chemical fat quantification. Correlation coefficients, intraclass correlation coefficients (ICC), receiver operating characteristic (ROC) analyses, and Bland-Altman analysis were performed to evaluate the relationships and agreement among the measurement methods. Results: Median values of ACM, MRI-PDFF, and laboratory fat content were 2.39 (1.87-2.84), 3.88 (2.46-7.60), and 1.24 (0.51-3.90), respectively. Strong correlations were observed between ACM and laboratory fat quantification (r = 0.878, p < 0.001) and between MRI-PDFF and laboratory analysis (r = 0.881, p < 0.001). ACM also correlated strongly with MRI-PDFF (r = 0.846, p < 0.001). The intraobserver reproducibility of ACM was excellent (ICC = 0.956, p < 0.001). AUROC values were 0.984 for ACM and 0.996 for MRI-PDFF, both indicating high diagnostic sensitivity and specificity. Conclusions: ACM demonstrated strong agreement with MRI-PDFF and chemical fat analysis in LFPs, supporting its potential as a reliable, accurate, and cost-effective non-invasive technique for hepatic steatosis quantification.