Evaluation of attenuation coefficient and sound speed for the diagnosis of liver steatosis: magnetic resonance imaging proton density fat fraction as a reference.

Abstract

Purpose: To compare the performance of attenuation coefficient (AC) and sound speed (SS) for evaluating liver steatosis in participants who underwent liver fat quantification using magnetic resonance imaging (MRI)-derived proton density fat fraction (PDFF).

Methods: Participants with chronic liver disease who underwent both MRI-PDFF (reference standard) and multiparametric ultrasound (US) measurements (such as for AC and SS) were retrospectively enrolled. The associations between the clinical variables and US markers were determined using multivariate linear regression. The diagnostic performance of US markers for detecting steatosis grades was evaluated using the area under the receiver operating characteristic curve (AUC).

Results: Fifty-three participants were included. Patients with metabolic dysfunction-associated steatotic liver disease were predominant (34 out of 53, 53.2%). Multivariate regression analysis revealed significant associations of AC with PDFF (regression coefficient: 0.05, p = 0.003) and body mass index (regression coefficient: 0.0094, p = 0.01). SS was also significantly associated with PDFF (regression coefficient: -12.20, P = 0.03) and body mass index (regression coefficient: -3.08, p = 0.01). The AUCs for AC and SS were 0.88 (95% CI: 0.79-0.97) and 0.70 (95% CI: 0.55-0.84), respectively, for identifying participants with steatosis grade S1 or higher. The performance of detecting steatosis grade S1 or higher was improved and more balanced when AC was ≥0.54 dB/cm/MHz and SS was ≤1536 m/s; the sensitivity and specificity were 72% and 86%, respectively.

Conclusion: SS was associated with PDFF and demonstrated fair diagnostic performance in identifying steatosis grade S1 or higher. The combined use of AC and SS may improve the detection of hepatic steatosis.