Singular value decomposition for physiological motion removal in acoustic-radiation-force-based cardiac shear wave elastography

Myocardial stiffness is a valuable biomarker aiding in cardiac diagnosis and can be measured with acoustic-radiation-force-based shear wave elastography (ARF-SWE). Although ARF-SWE acquisitions are short (order of 10 ms), rapid cardiac motion can still corrupt the measurements, with magnitudes significantly larger than the ARF-induced particle displacements. Singular value decomposition (SVD) has demonstrated its effectiveness in clutter filtering for ultrasound flow measurements. Here, we apply SVD for physiological motion suppression in cardiac ARF-SWE. By performing SVD filtering on SWE data acquired from an ex-vivo porcine heart beating in Langendorff mode, we show that the SVD filter outperformed the existing band-pass filter and average wall motion subtraction method both qualitatively and quantitatively in terms of the filtered ARF-induced particle motion and downstream shear wave speed.