Objective: To assess the effects of frame and volume rate on the concordance between two-dimensional speckle tracking strain (2DS) and three-dimensional speckle tracking strain (3DS), and between 2DS and triplane imaging of speckle tracking (Tri-P).
Methods: Global longitudinal strains (GLSs) derived from 2DS, 3DS, and Tri-P were compared among 142 prospectively recruited patients who underwent evaluation of subclinical left ventricle (LV) function.
Results: Feasibility to obtain GLS of 3DS was significantly higher than that of Tri-P (76% vs. 47%, P < 0.001). The correlation between 2DS and 3DS was only modest (r = 0.47) whereas that of 2DS and Tri-P was better (r = 0.67). The difference in frame/volume rate between two methods also affected their correlation. A volume rate between 34 and 50 volumes/sec had the highest correlation between 2DS and 3DS (r = 0.72). The correlation between 2DS and Tri-P was better with a difference in frame-rate ≤20 per second than with a difference >20 per second. Likewise, there was a better correlation between 2DS and 3DS when the difference between 2D frame rate and 3D volume rate was ≤40 per second, compared to when it was >40 per second. These associations differed from segment to segment and the apical segments had the highest correlation and the basal the lowest.
Conclusions: The feasibility of each means of strain calculation showed important differences, with 2DS being the most attainable. Strain values were not interchangeable among 2DS, Tri-P, and 3DS. Importantly, poor correlations seemed to be driven by differences in acquisition rate. Currently, 2DS offers the most robust measurement of subclinical myocardial dysfunction.
©2011, Wiley Periodicals, Inc.