Comparison of different approaches to assess fascicle length using 3D ultrasound with cadaveric dissection as a reference for the human biceps femoris long head

Purpose The biceps femoris long head (BFlh) has a complex architecture, posing challenges for accurately assessing its fascicle length. This study compared four 3D ultrasound (3DUS)-based methods for estimating mid-belly BFlh fascicle length in human cadavers (n = 6 legs). Methods Method 1 and Method 2 assumed straight fascicles within a single longitudinal plane, aligned with the muscle axis and oriented perpendicular to either proximal or distal aponeurosis. Method 3 represented multiplanar trajectory using two connected straight segments, whereas Method 4 further incorporated local in-plane curvature within each segment. Results Agreement relative to the direct measurement of fascicle length by dissection using Bland–Altman bias and 95% limits of agreement (LoA), together with mean difference. All 3DUS-based estimates showed good agreement with dissection (bias < 7.5%, LoA < 18.8 mm, mean differences < 7 mm). Method 3 and Method 4 yielded the lowest bias, narrowest limits of agreement, and smallest mean differences. While repeated-measures ANOVA revealed a significant main effect, post hoc analysis did not identify significant differences between methods (all p ≥ 0.096). All methods demonstrated good intra-rater repeatability (ICC (3,1) = 0.77–0.87) and moderate measurement precision (SEM = 5.51–8.74%). Conclusion Overall, these findings suggest that a two-plane approach yields higher accuracy for BFlh fascicle length measurements than approaches assuming straight fascicles within a single plane.