Layer-Specific Radiofrequency Ultrasound-Based Strain Analysis in a Porcine Model of Ischemic Cardiomyopathy Validated by a Geometric Model

F.J. van Slochteren, T.I.G. van der Spoel, H.H.G. Hansen, P.H.M. Bovendeerd, P.A. Doevendans, J.P.G. Sluijter, S.A.J. Chamuleau and C.L. de Korte

Ultrasound in Medicine and Biology 2014;40:378-388

DOI PMID

Abstract

Local layer-specific myocardial deformation after myocardial infarction (MI) has not been studied extensively although the sub-endocardium is more vulnerable to ischemia and interstitial fibrosis deposition. Radiofrequency (RF) ultrasound-based analysis could provide superior layer-specific radial strain estimation compared with clinically available deformation imaging techniques. In this study, we used RF-based myocardial deformation measurements to investigate layer-specific differences between healthy and damaged myocardium in a porcine model of chronic MI. RF data were acquired epicardially in healthy (n = 21) and infarcted (n = 5) regions of a porcine chronic MI model 12 wk post-MI. Radial and longitudinal strains were estimated in the sub-endocardial, mid-wall and sub-epicardial layers of the left ventricle. Collagen content was quantified in three layers of healthy and infarcted regions in five pigs. An analytical geometric model of the left ventricle was used to theoretically underpin the radial deformation estimated in different myocardial layers. Means ± standard errors of the peak radial and longitudinal strain estimates of the sub-endocardial, mid-wall and sub-epicardial layers of the healthy and infarcted tissue were: 82.7 ± 5.2% versus 39.9 ± 10.8% (p = 0.002), 63.6 ± 3.3% versus 38.8 ± 7.7% (p = 0.004) and 34.3 ± 3.0% versus 35.1 ± 5.2% (p = 0.9), respectively. The radial strain gradient between the sub-endocardium and the sub-epicardium had decreased 12 wk after MI, and histologic examination revealed the greatest increases in collagen in the sub-endocardial and mid-wall layers. Comparable normal peak radial strain values were found by geometric modeling when input values were derived from the in vivo measurements and literature. In conclusion, the estimated strain values are realistic and indicate that sub-endocardial radial strain in healthy tissue can amount to 80 This high value can be explained by the cardiac geometry, as was illustrated by geometric modeling. After MI, strain values were decreased and collagen content was increased in the sub-endocardial and mid-wall layers. Layer-specific peak radial strain can be assessed by RF strain estimation and clearly differs between healthy and infarcted tissue. Although the relationship between tissue stiffness and tissue strain is not strictly local, this novel technique provides a valuable way to assess layer-specific regional cardiac function in a variety of myocardial diseases.