Purpose
The aim of the study was to test inter‐observer and intra‐observer reliability of measuring multiparametric ultrasound in adult livers.
Methods
We prospectively measured shear wave velocity (SWV, m/s), shear wave dispersion slope (SWD, m/s/kHz), attenuation coefficient (ATI, dB/cm/MHz), normalized local variance (NLV), and echo intensity ratio of liver to kidney (L/K ratio) in 21 adults who underwent liver magnetic resonance imaging‐proton density fat fraction (MRI‐PDFF). Intraclass correlation coefficient and 95% Bland–Altman limits of agreement (95% LOA) were used to analyze intra‐ and inter‐observer reproducibility.
Results
Based on liver MRI‐PDFF, 21 participants (8 men and 13 women, mean age 55 years) were divided into group 1 (11 normal livers, MRI‐PDFF <5%) and group 2 (10 steatotic livers, MRI‐PDFF ≥5%). ICCs for intra‐observer repeatability and inter‐observer reproducibility in measuring multiple ultrasound parameters in both normal and steatotic livers were above 0.75. However, 95% confidence interval for measuring SWD in all livers and L/K ratio in normal livers was 0.38–0.90 and 0.47–0.91, respectively. Differences in SWV, SWD, ATI, NLV, L/K ratio, and MRI‐PDFF between participants with and without hepatic steatosis were significant (p < .05), whereas serum biomarkers and body mass index were not (p > .05), based on a two‐tailed t‐test.
Conclusions
The results suggest that the repeatability and reproducibility for measuring liver SWV, ATI, and NLV are moderate to excellent, while those for SWD and L/K ratio are poor. Standardized machine settings, scanning protocols, and operator training are suggested in performing multiparametric ultrasound of the liver.