Intra-observer and device-dependent inter-observer reliability of contrast-enhanced ultrasound for muscle perfusion quantification

Muscle perfusion quantification by contrast-enhanced ultrasound (CEUS) may facilitate treatment decisions in musculoskeletal disorders. Translation into clinical routine relies on high intra-observer and interobserver reliability and transferability between ultrasound devices to enable validation and multicenter studies. This study evaluates these aspects for deltoid muscle perfusion quantification, including possible multicenter study setups. One hundred sixty-six CEUS quantifications were conducted on 42 shoulders. Intra-observer reliability revealed a high intra-class correlation coefficient (ICC, r = 0.91) and low coefficient of variation (CV, 10.28%). Inter-observer reliability revealed an ICC of .84 and a CV of 17.1%, but these values decreased when different ultrasound devices were used (ICC = .60, CV = 18.6%). Re-evaluating subgroups with high sectional plane concordance significantly increased reliability (intra-observer: ICC = .97, CV = 5.49%, inter-observer/same device: ICC = .98, CV = 5.83%, varying devices: ICC = .78, CV = 9.8%). CEUS perfusion quantification of the deltoid seems applicable for multicenter studies, yet pooling different ultrasound devices remains critical. Sectional plane concordance appears to be crucial for reliability and transferability of CEUS muscle perfusion quantifications. (C) 2019 World Federation for Ultrasound in Medicine & Biology. All rights reserved.