Assessment of Lower Limb Muscle Volume Using 3D Ultrasonography: Validity and Reliability Compared to MRI

Muscle volume is a key indicator of muscle strength and neuromuscular health, traditionally assessed using Magnetic Resonance Imaging (MRI). While accurate, MRI is costly and time-intensive. Three-dimensional ultrasonography (3DUS) offers a more accessible alternative but requires validation due to its setup-dependent accuracy. Here, we investigate the validity and reliability of a custom 3DUS setup for measuring lower limb muscle volumes. We compared the 3DUS assessments to MRI, and documented the methodology for broader research adoption. Fifteen participants (7 males, 8 females; 18–40 years) underwent two 3DUS sessions and one MRI (MAGNETOM Prisma, Siemens, Erlangen, Germany) session. Target muscles— m. tibialis anterior, m. vastus lateralis, mm. gastrocnemii and m. biceps femoris —were scanned using ultrasonography (Acuson Juniper, Siemens, Erlangen, Germany) integrated with a motion capture system (Optitrack, NaturalPoint, Corvallis, OR, USA). Phantom models (100–600 mL) were scanned to enhance validation. Subsequent to scanning ten participants (6 males and 4 females), we adapted our scanning protocol. 3DUS volumes were analyzed in 3D Slicer by two raters while MRI images were analyzed using Dafne. Reliability metrics included intra-class correlation (ICC), coefficients of variation (CV%), standard error of measurement (SEM), and minimal detectable change (MDC). 3DUS demonstrated excellent test-retest and inter-rater reliability, with ICCs ranging from 0.97 to 0.99 and CV% values between 2.0–4.6%. MDC values were below 5 mL (0.6-2.7%) for all muscles. However, 3DUS systematically underestimated muscle volumes compared to MRI, with biases ranging from -15.4% to -44.6%. The highest agreement was found for the tibialis anterior, the lowest for the gastrocnemii muscles. When adapting the scanning protocol, the mean difference between the 3DUS and MRI measured volumes were reduced by about 70%. Phantom scans confirmed high accuracy for both modalities, suggesting in vivo errors stemmed from probe pressure and sweep inconsistencies. 3DUS showed excellent reliability but limited comparability to MRI, with underestimation influenced by muscle shape and location. Systematic underestimation compared to MRI underscores the need for methodological refinements. Despite limitations, 3DUS is a promising, cost-effective tool for longitudinal muscle assessments, especially for detecting intervention-induced changes. Open documentation of our methodology supports reproducibility and further adaptation and innovation.