Compound strain imaging for elastic modulus reconstruction of transverse vessel cross-sections

Atherosclerotic plaque rupture can initiate stroke or myocardial infarction. Lipid-rich plaques with a thin fibrous cap are more prone to rupture than fibrotic plaques. The elastic Young's modulus differs for the various plaque components. The goals of this study were 1) to reconstruct Young modulus images for transverse vessel cross-sections based on 2D displacement fields of the vessel wall obtained noninvasively with a linear array transducer and 2) to investigate if the reconstruction is improved when using multi-angle displacement compounding instead of single angle displacement estimation. Simulated and experimental radiofrequency echo data were generated for beam steering angles of -30°, 0° and 30° for three vessel-mimicking phantoms: a homogeneous phantom with a concentric lumen, a homogeneous phantom with an eccentric lumen and a two-layered phantom with a soft layer inside and an eccentric lumen. Displacement fields were estimated using a coarse-to-fine 2D cross-correlation based algorithm. Relative Young's modulus images were reconstructed using either compounded displacement fields or 0°angle displacement fields. To determine the performance of the modulus reconstruction the median absolute differences between the measured and model predicted axial and lateral displacements were calculated. Furthermore, the median and inter-quartile range (IQR) of the relative modulus estimates for each separate phantom layer were calculated and compared to the real values. The median difference between lateral displacements used as input for the reconstruction and those corresponding to the reconstructed modulus image reduced twofold to threefold when using multi-angle compounding instead of single angle imaging. This also resulted in improved Young's modulus images: the IQR of the relative Young's modulus reduced approximately a factor 2 when using compounding. The errors in relative Young's modulus with compounding were maximally ~10%. Thus, more accurate modulus reconstructions for transverse cross-sections of vascular structures can be obtained with compounding than without compounding. Furthermore, evaluation using in vivo data is required to demonstrate the clinical benefit of this technology for vulnerable plaque detection.