An investigation into the variability between different shear wave elastography systems in muscle

Aims: The reliability and agreement between shear wave elastography (SWE) systems using different acquisition methods in muscles is not yet established. The objectives were to determine the reliability of a new SWE system on normal resting muscles using different acquisition methods and to compare its performance to an established state-of-the-art system. Material and methods: Small, medium and large ROI sizes in addition to longitudinal, oblique and transverse orientations over five different locations within the rectus femoris muscle were tested using the new system. Results were compared to the established system to test for inter-system reproducibility. Results: Lowest within-subject coefficient of variance (4.3%) and shear wave velocity (1.83 m/s) were associated with the medium ROI and longitudinal orientation from the lateral location. This combination resulted in a strong internal agreement of intra-class correlation of 0.76 (0.57-0.89) for the new system and an almost perfect agreement of 0.92 (0.82-0.97) for the established. Inter-system reproducibility for the best combination was 0.71 (0.48-1) with a mean velocity difference +/- 95% limits of agreement of 0.07 +/- 0.49 m/s. Conclusions: Altering SWE acquisition methods can produce variable results. The new system produced reliable results that are comparable with but not as reliable as the established.