Ultrasound 2D-SWE inter-system variability and reliability: Insights from an elasticity phantom model

Two-dimensional shear wave elastography (2D-SWE) enables quantitative assessment of tissue stiffness. However, due to system-specific variability in shear wave generation, cross-platform reproducibility needs to be assessed. Therefore, this study aimed to assess the inter-system agreement of 2D-SWE measurements between two widely used ultrasound systems (Philips EPIQ and GE LOGIQ) using a standardized tissue-mimicking elasticity phantom with known stiffness values. A total of 320 elasticity measurements, expressed in kilopascals (kPa), were obtained from four lesions representing varying stiffness levels. A total of 160 measurements were obtained per system, comprising 40 measurements for each lesion. Inter-system agreement of 2D-SWE measurements was asssessed using the intraclass correlation coefficient (ICC) and Bland-Altman plot, and the differences in 2D-SWE measurements between the two ultrasound systems were assessed using the Wilcoxon-rank test. Significant differences were observed in stiffness measurements. For Lesion I (reference: 8 ± 3 kPa), the EPIQ overestimated stiffness (9.72 ± 0.93 kPa), while the LOGIQ closely aligned with the reference (7.95 ± 0.44 kPa). Both systems underestimated high stiffness levels, such as Lesion IV (reference: 80 ± 8 kPa), reporting 11.23 ± 2.04 kPa (EPIQ) and 20.33 ± 2.64 kPa (LOGIQ). The ICC for inter-system agreement was 0.81 (p < 0.0001), indicating good overall consistency. Bias in measurements was 0.78 ± 6.45 kPa. There were significant differences in the 2D-SWE measurements for all lesions between the two ultrasound systems (p < 0.0001). These findings highlight the need for standardization and calibration protocols to address cross-platform inconsistencies. Future research should focus on validating findings in clinical settings and exploring solutions to improve reproducibility and reliability across systems.