Ultrasound and MRI Are Not Interchangeable for Measuring Muscle Cross-Sectional Area: Analysis of Reliability and...

Ultrasound-based anatomical cross-sectional area (US-ACSA) measurement using extended-field-of-view (EFOV) imaging offers a practical alternative to magnetic resonance imaging (MRI-ASCA) measurements. Although its reliability and validity have previously been assessed, previous studies have focused only on a single or limited number of muscles with little attention to effects of muscle size and sex. This study was conducted in ten lower limb muscles of varying sizes in males and females in order to assess the reliability and validity of US-ACSA measurements using MRI as the gold standard method. Twelve males and twelve females participated in this study. US- and MRI-ACSA were measured twice, one day apart. Reliability was assessed using intraclass correlation coefficients (ICC_2,3), standard error of measurement (SEM) and minimal detectable change (MDC). Validity was examined using Pearson product-moment correlation and Bland-Altman analysis. A linear model was used to evaluate the effects of muscle size and sex. ICC_2,3, SEM and MDC ranged from 0.992% to 0.999%, 0.50%-1.90% and 1.25%-5.36% for US-ACSA and from 0.952% to 0.998%, 0.90%-4.76% and 2.51%-13.19% for MRI-ACSA. US-ACSA showed strong correlations with MRI-ACSA (r = 0.73-0.95). A systematic underestimation of MRI values (relative bias: -11.5% to -23.7%) was identified, with no systematic effect of muscle size or sex. These findings suggest that US-ACSA measurements of lower limb muscles are highly reliable regardless of muscle size and sex. However, due to the large systematic bias, US- and MRI-based ACSAs measurements are not interchangeable.