Inter-operator and inter-device reproducibility of shear wave elastography in healthy muscle tissues.

Abstract

PurposeThe study aimed to assess whether the more limiting factor in reproducibility of shear wave elastography (SWE) would be the operator dependency or the incompatibility of different ultrasound (US) devices. The interrater agreement with less experienced operators was studied.MethodsA total of 24 healthy volunteers participated in the study (18 females, 6 males; range of age 27–55 years). SWE of biceps brachii (BB) and tibialis anterior (TA) muscles was performed on both sides from all participants in both longitudinal and transverse orientation of the transducer in respect to muscle fibers. Two operators repeated the SWE with two different US devices from different manufacturers (scanners 1 and 2).ResultsIntraclass correlation coefficient between the two operators was 0.91 (CI 0.88–0.93) for scanner 1 and 0.81 (CI 0.74–0.86) for scanner 2, respectively. Instead, there were significant differences in the SWE measurements between the two scanners, emphasizing in transverse orientation of the transducer. In the transverse transducer orientation, the mean shear wave velocity (SWV) in TA was 1.45 m/s (standard deviation [SD] ± 0.35 m/s) with scanner 1 and 2.35 m/s (SD ± 0.83 m/s) with scanner 2 (p < 0.001). In BB, the mean transverse SWV was 1.49 m/s (SD ± 0.35 m/s) with scanner 1 and 2.29 m/s (SD ± 0.63 m/s) with scanner 2 (p < 0.001). In longitudinal transducer orientation, the mean SWV in TA was 3.00 m/s (SD ± 0.73 m/s) with scanner 1 and 3.26 m/s (SD ± 0.42 m/s) with scanner 2 (p = 0.050). In BB, the mean longitudinal SWV was 3.60 m/s (SD ± 0.77 m/s) with scanner 1 and 3.96 m/s (SD ± 0.62 m/s) with scanner 2 (p = 0.019). The presented mean values were obtained by operator 1, there were no significant differences in the SWE measurements performed by the two operators.ConclusionThe results implicate that the reproducibility of the SWE measurements depends rather on the used US device than on the operator. It is recommendable that clinics collect reference values with their own US device and consider threshold values presented in previous studies only directional.