In vivo validation of 2D myocardial elastography at variable levels of ischemia

Abstract

The objective of this study is to validate two-dimensional (2D) myocardial elastography estimates in canine hearts in vivo against direct sonomicrometry measurements at variable levels of myocardial ischemia induced by the occlusion of the left anterior descending (LAD) coronary artery. Two-dimensional myocardial elastography comprised transmural: 1) 2D (lateral and axial) displacements using 1D cross-correlation in a 2D search and recorrelation (3.5 mm window size; 80% overlap); 2) 2D Lagrangian strain estimation using a least-squared strain estimator; and 3) polar (radial and circumferential) strains through coordinate transformation. An Ultrasonix RP system with a 3.3 MHz phased array and 128 beams was used to acquire RF frames in a 2D short-axis view at the frame rate of 211 fps using an automatic composite technique previously developed by our group. A non-survival canine ischemic model was performed by inducing an occlusion of the LAD coronary artery, from 0% to 100% at a 20% increment. Two separate sets of four piezoelectric crystals (1 in the sub-epicardium; 1 in the sub-endocardium; 2 at the epicardium) in a tetrahedral configuration were placed in the ischemic (i.e., anterior) and normal (i.e., posterior) regions. Polar strains were calculated from the coordinates of the crystals. Reduced radial thickening or radial thinning at the occlusion level larger than 40% were observed based on strain estimates from both 2D myocardial elastography and sonomicrometry. Good agreement in radial strain between the two methods was found from the Bland-Altman analysis. These findings demonstrate that 2D myocardial elastography could serve as a powerful diagnostic tool to non-invasively detect, localize and identify the ischemic region at its very early stages, i.e., at 40% coronary artery occlusion.