Performance of elastic image registration against speckle tracking for 2D cardiac motion and strain estimation

Abstract

Despite the availability of multiple solutions for assessing myocardial strain, little is currently known about the relative performance of the different methods. In this study, we sought to contrast two popular strain estimation techniques directly (block-matching and registration) in an in-vivo setting by comparing both to a gold standard reference measurement. Mid short-axis images were acquired with a GE Vivid7 ultrasound system in five open-chest sheep instrumented with ultrasonic microcrystals. Radial (epsilon(RR)) and circumferential strain (epsilon(CC)) were estimated during four stages: at rest, during esmolol and dobutamine infusion, and during acute ischemia. While the correlations of a well-validated block matching approach and the elastic registration method were comparable for the epsilon(CC) component (0.82 vs 0.81 respectively), imposing regularization during the motion estimation process (i.e. as embedded in the elastic registration method), showed to be advantageous to estimate epsilon(RR) as it improved performance significantly (0.63 vs 0.83; p=0.05). Moreover, the bias and the limits of agreement were also smaller in this direction (p<0.01).