Consistency and reliability of ultrasound-derived fat fraction in hepatic steatosis assessment: influence of posture and breathing variations

Given the recent redefinition of metabolic dysfunction-associated steatotic liver disease (MASLD) and the clinical demand for non-invasive tools, this study evaluated the effect of body position (e.g., supine, left lateral recumbent) and respiratory status (e.g., deep inspiration, expiration) on ultrasound-derived fat fraction (UDFF) measurements., assessed their reliability and consistency, and determined the optimal conditions for obtaining UDFF. A retrospective analysis was performed using the system with the UDFF algorithm. Two operators performed UDFF measurements in six different scenarios, each consisting of three measurements (18 in total). Consistency analysis was achieved primarily by intragroup correlation coefficients (ICC) and Friedman’s test, and Spearman’s correlation analysis mainly achieved correlation. 319 participants were included (mean age 40 ± 17 years; 154 males). The ICC showed strong reliability (mean ICC = 0.999), and Friedman’s test showed no significant differences between BMI groups or between the various breathing and postural conditions (p > 0.05). However, the coefficient of variation (Cov), which affects the results, was lowest for supine breath-holding. Spearman’s ρ showed a strong correlation between supine breath-holding (ρ = 0.973) and end-inspiratory and end-expiratory breathing (ρ = 0.965). Body mass index was also closely and positively correlated with UDFF (ρ = 0.652; p < 0.001), and highly correlated across positions and respiratory states (ρ = 0.954–0.973; p < 0.001). UDFF demonstrates high reliability and consistency, particularly in the supine breath-hold state, offering a standardized protocol for patients with limited mobility or respiratory distress, such as those with COPD or pediatric populations. In addition, the supine breath-hold state may be a more appropriate measurement condition.