Regional correlations between MRI-Derived muscle cross-sectional area and ultrasound-derived muscle thickness of the Vastus Lateralis

Abstract

Problem statement and purpose: Magnetic resonance imaging (MRI) has been regarded as the gold standard for quantifying vastus lateralis (VL) muscle size; however, its substantial cost has restricted its accessibility in resource-limited environments. Ultrasonography (US) has emerged as an affordable alternative. However, the relationship between regional VL muscle size measurements obtained via MRI and US remains uncertain. Consequently, this study aimed to investigate the correlations between the vastus lateralis cross-sectional area (VLCSA) derived from MRI and the vastus lateralis muscle thickness (VLMT) derived from US across three regions: proximal, middle, and distal. Methods: Eighteen healthy, physically active undergraduate students (age = 21.1 ± 0.5 years; height = 174.1 ± 5.3 cm; body weight = 68.8 ± 11.3 kg) underwent both US and MRI measurements of the VL muscle on the same day in a randomized order. Images of VLCSA and VLMT were acquired at the proximal, middle, and distal regions of the muscle. One-way ANOVA was used to compare VLCSA and VLMT among these regions. Pearson's correlation coefficient analyses were performed to examine the relationships between VLCSA and VLMT at each corresponding region and the sum of the three regions. Results: Statistical analyses revealed significant differences among regions for both VLCSA (F[2,51] = 24.616, p = 0.000) and VLMT (F[2,51] = 7.786, p = 0.001). Notably, a strong and significant correlation between VLCSA and VLMT was observed only in the middle region (r = 0.72, p = 0.001), while weaker and non-significant correlations were found at the proximal (r = 0.35) and distal regions (r = 0.45). However, when the VLMT and VLCSA values from all three regions were summed, a moderate and significant correlation was observed (r = 0.54, p = 0.02). Conclusion: Our findings suggest that prioritizing the middle region for US measurement of VLMT is advisable because this region exhibits a stronger correlation with MRI results. Furthermore, we recommend that practitioners assess VLMT across multiple regions rather than relying on a single, randomly selected region. By summing the measurements from multiple regions, a more reliable estimation of VLCSA can be obtained compared to using a single region alone.