Contents I. Introduction
Myocardial strain is measured as an important indicator to quantify myocardial function [1]. Doppler-based methods are traditionally used to measure myocardial strain [2]. A major disadvantage of this approach is the fact that it is only capable of measuring strain along the direction of propagation of the ultrasound wave, yielding a one dimensional strain measurement [3]. Methods based on tracking B-mode patterns or radio-frequency patterns within the image plane allow to measure myocardial strain in two dimensions [4]. Although this is obviously an improvement, it remains limited as cardiac motion and deformation is truly 3D. With the introduction of matrix array transducer technology, 3D ultrasound (US) imaging of the heart has become feasible [5]. However, the increased field of view in 3D US comes at the expense of a reduction in both spatial and temporal resolution, making decorrelation between consecutive image frames significant and therefore challenging 3D strain measurements. A prior study of our group showed that spatio-temporal elastic registration of such 3D volumetric ultrasound data sets can enable the measurement of the full 3D strain tensor in-vivo [6].
