Inter-System Comparison of Shear Wave Elasticity Imaging in Patients Undergoing Bariatric Surgery
H Whitson1*, A Possell1, J Zagzebski1, C Sirlin2, S Reeder1, T Hall1, I Rosado-Mendez1, L Mankowski-Gettle1, (1) University of Wisconsin, Madison, WI, (2) University Of California - San Diego
Presentations
(Saturday, 3/26/2022) 10:30 AM - 12:30 PM [Central Time (GMT-5)]
Room: Celestin D-E
Purpose: Ultrasound-based shear-wave elastography imaging estimates the shear wave speed (SWS), a recommended biomarker for liver fibrosis. However, SWS bias and variability increase with tissue depth, an important challenge in patients with obesity, who are at elevated risk for liver fibrosis. This study presents a preliminary comparison of SWS values obtained from three imaging systems in phantoms and in vivo from adults undergoing bariatric surgery.
Methods: Phantom data was collected following the Quantitative Imaging Biomarkers Alliance protocol for assessing bias and precision for SWS estimates. Two phantoms, denoted L and H for low and high stiffness, were measured at two depths with three ultrasound systems, denoted systems A-C. Each phantom data set was assessed using within-subject coefficient of variation (wCV) and percent bias from the overall mean. Systems A-C were then used to compare SWS in the livers of thirty-five adult patients with obesity. Pearson correlation coefficients were computed between median SWS values of the systems and discrepancies between systems were compared to BMI and depth.
Results: In phantoms, the maximum percent bias and wCV in phantoms were -4.13% and 2.91% respectively. In adults with obesity, the median (IQR) values of SWS were 1.27 (1.11-1.51) m/s, 1.03 (0.96-1.15) m/s, and 1.67 (1.39-1.91) m/s for systems A, B, and C respectively. Discrepancies between pairs of systems yielded median (IQR) values of 0.18 (0.10-0.48) m/s, 0.49 (0.19- 0.74) m/s, and 0.53 (0.36- 0.86) m/s for system discrepancies A/B, A/C, and B/C. Pearson correlation coefficients between each system pair were 0.160, 0.011, and 0.303 for combinations A/B, A/C, and B/C.
Conclusion: Phantom results were reproducible at two depths in phantoms of different stiffness. However, in adults with obesity, SWS measurements correlated poorly between systems. Further research is needed to confirm our results and assess additional sources of variability in our protocol.
Funding Support, Disclosures, and Conflict of Interest: The work is funded through the National Institutes of Health grant R01 DK088925. Additionally, UW-Madison receives research support from GE Healthcare and Siemens Healthineers.
Keywords
Taxonomy
Contact Email