Ultrasound-Based Speckle Tracking For 3d Strain Estimation Of The Arterial Wall - An Experimental Validation Study In A Tissue Mimicking Phantom

Arterial stiffness is an important risk factor for cardiovascular disease. As such, ultrasound-based methods have been proposed to assess arterial strain as a measure of stiffness. The aim of the current study was to validate our recently proposed speckle tracking (ST) algorithm to estimate the in-plane wall strain tensor in an experimental setup. Three polyvinyl alcohol phantoms mimicking the carotid artery were constructed with different mechanical properties (2, 3 and 4 freeze-thaw cycles). The phantoms were connected to a pump, programmed to generate carotid flow profiles at peak flows of 7, 14, 21, 28 and 35 ml/s. Long and short-axis ultrasound images were obtained using a Vivid7 Dimension system. Radial, longitudinal and circumferential strains were estimated using the ST algorithm (kernel size: 2.7 lambda x2 lambda, normalized cross-correlation; spline inter-polation for subsample motion estimation; 40% window overlap). Sonomicrometry was used to acquire reference values of strain in the phantoms. Good agreement was found between the estimated radial, longitudinal and circumferential strain and the acquired reference strain. The correlation between estimated mean peak strain values and reference peak strain values was r = 0.92 (p < 0.001) for radial strain, r = 0.72 (p = 0.006) for longitudinal strain and r = 0.91 (p < 0.001) for circumferential strain.