An optimisation-based iterative approach for speckle tracking echocardiography

Abstract

Speckle tracking is the most prominent technique used to estimate the regional movement of the heart based on echocardiograms. In this study, we propose an optimised-based block matching algorithm to perform speckle tracking iteratively. The proposed technique was evaluated using a publicly available synthetic echocardiographic dataset with known ground-truth from several major vendors and for healthy/ischaemic cases. The results were compared with the results from the classic (standard) two-dimensional block matching. The proposed method presented an average displacement error of 0.57 pixels, while classic block matching provided an average error of 1.15 pixels. When estimating the segmental/regional longitudinal strain in healthy cases, the proposed method, with an average of 0.32 ± 0.53, outperformed the classic counterpart, with an average of 3.43 ± 2.84. A similar superior performance was observed in ischaemic cases. This method does not require any additional ad hoc filtering process. Therefore, it can potentially help to reduce the variability in the strain measurements caused by various post-processing techniques applied by different implementations of the speckle tracking. Graphical Abstract Standard block matching versus proposed iterative block matching approach.

Keywords: Echocardiography; Myocardial deformation; Speckle tracking echocardiography; Strain imaging.

Graphical Abstract

Standard block matching versus proposed iterative block matching approach

Fig. 1

Speckle tracking using BM where a region in the image (kernel) is selected and sought for in the next image by sequentially trying out different positions, testing the similarity between the kernel and the pattern observed in that position. The position where the similarity between the kernel and the observed pattern is maximal is accepted as the new position of the original kernel

Fig. 2

Flowchart showing the steps involved in solving the proposed optimisation-based tracking algorithm

Fig. 3

An example A4C from the Siemens healthy sequence and corresponding displacement vector fields during the rapid ejection phase (peak systole). a Zoomed view of LV cropped from the original image. b Ground-truth. c–d Displacement fields obtained from standard BM and optimised BM approach in the rapid ejection phase, respectively. Corresponding figures for other vendors are provided in Appendix A

Fig. 4

Boxplots of the error for the healthy sequence from Siemens. The error is computed as the magnitude of the difference between the calculated and ground-truth displacement vectors and is provided for standard (left) and proposed (right) tracking methods. The x-axis shows the frame number

Fig. 5

Displacement error for the healthy A4C synthetic sequences across all vendors for the two speckle tracking approaches. The horizontal line represents mean; the box signifies the quartiles, and the whiskers represent the 2.5% and 97.5% percentiles

Fig. 6

Top: an A4C view with the LV myocardium segmentation regions overlaid. Below: violin plots of the error in the segmental strain measurements for the healthy synthetic sequence from Siemens. The solid black line represents mean, and the green line represents the median; the box signifies the quartiles, and the whiskers represent the 2.5% and 97.5% percentiles

Fig. 7

Comparison of GLS measurements obtained from the standard BM approach for the healthy and ischaemic LCX (ischaemic left circumflex coronary artery) cases across all vendors with the known ground-truth. The solid and dashed blue lines represent the calculated strain values for healthy and ischaemic cases, respectively. The solid and dashed magenta lines indicate the corresponding ground-truth

Fig. 8

Same as Fig. 7, but for the optimised BM approach

Fig. 9

Example of a (presumably) “common sense” editing in a commercial package on one frame, where 3 regional strain values are given for the lateral wall, but there is no visible myocardium to be tracked in that region