5C-5 A Rigid Wall Approach to Physiologic Motion Rejection in Arterial Radiation Force Imaging

Abstract

Physiologic motion corrupts measurements of induced tissue displacements and obscures tissue mechanical properties in radiation force ultrasound. Wall dilation and contraction with cardiac pulsation is especially disruptive to radiation force imaging the arterial system. We hypothesize that exploiting a rigid arterial wall model, which assumes long wavelength arterial pulse waves, will improve physiologic motion rejection in arterial radiation force imaging. Three rigid wall assuming filters (polynomial regression, principal component regression, and FIR high-pass filters) were compared to four filters that did not assume a rigid arterial wall (linear regression, quadratic regression, principal component regression, and FIR high-pass filters). The filters were tested using Field II generated data inclusive of simulated arterial wall motion combined with experimental acoustic radiation force impulse (ARFI) or shear wave elastography imaging (SWEI) displacement profiles. Performance metrics were sum of absolute differences (SAD) between original and filtered ARFI or SWEI displacement profiles in terms of total profile error, measured peak displacement error, measured recovery time error, and time-to-peak displacement error. Rigid wall assuming polynomial and principal component regression filters yielded the lowest SAD scores. The filters were also qualitatively compared on in vivo ARFI and SWEI data acquired in healthy pig iliac arteries.