Measurements of transmural strain variations by two dimensional ultrasound speckle tracking

Abstract

Background: Myocardial infarction (MI) is known to progress from the inner layers towards the epicardium. Since it is important to detect MI early, to prevent a negative remodeling process of the left ventricle (LV), the hypothesis of this study was that evaluation of layer-specific strains is feasible, and it will enable differentiation between subjects with large MI, small MI and normal LV. Methods:In this study a commercial speckle tracking echocardiography (STE) program was modified to measure the strains at three myocardial layers instead of for the total-wall-thickness. After a validation process by using software implemented phantoms, the commercial and modified programs were applied to echocardiography of 54 subjects. Results:The validation study results for 972 segments showed an agreement between the endocardial strains, manually measured by the commercial program, and automatically measured by the modified program. Finally, the algorithm was applied to scans of 15 normal subjects, 9 patients with small MI and to 6 patients with large MI. The results show that the strain elevated from the endocardium towards the epicardium for the normal and small MI groups, but not for the large MI group. Conclusions:In conclusion, the layer-specific STE method allows accurate analysis of the transmural variations of the strains.

Feasibility of reproducible vendor independent estimation of cardiac function based on first generation speckle tracking echocardiography

“…The EWST algorithm is primarily based on the method previously reported [23,26,27] . The algorithm contains a single smoothing/weighting function, with one parameter N, which controls the amount of spatial smoothing.…”

“…Three chains of points, knots [23] , are defined within the ROI, one imposed along the inner (subendocardial) boundary of the myocardium, one along the outer (subepicardial) boundary, and one along the line in between (mid wall). The 3 chains of knots are assumed to be along 3 smooth curves and are further assumed to mark the same physical points at all frames.…”

“…First, the displacement field of the TP is smoothed using a weighted bilinear approximation. In the second stage, the displacements of all the knots, along the 3 chains are smoothed, using a smoothing spline, as detailed elsewhere [23] . The open-source STE algorithm proposed here allows the operator, differently from previously reported algorithms, a control of the degree of the smoothing of the spline by defining a weighting-parameter N, which will effectively increase the amount of smoothing for noisy data.…”

“…Thus, though the Strain Standardization Initiative reduced significantly the inter-vendor variability of the strain measurements, a consequent study [14] concludes that current vendor independent STE software do and the sam Speckle tracking is mostly performed by block-matching [19][20][21][22] , where the displacement of a material point is estimated by searching for the "best match" between a block in the source frame and a block in the target frame. The center of the block in the source frame is usually defined as a "tracking point" (TP) [23] . The displacement search is usually limited to a region of search (ROS) around the TP, based on an estimation of the maximal tissue velocities.…”