|| Checking for direct PDF access through Ovid
Aim: Analysis of the agreement between 3D area strain and 3D/2D longitudinal strain in the assessment of both normal and pathological LV wall motion patterns and of its vulnerability to image and tracking artifacts.Methods: 25 patients with normal regional LV function and 7 patients with anterior myocardial infarction underwent 2D and 3D echocardiography, followed by offline speckle tracking measurements. Systolic strain patterns of 3D area and 3D longitudinal strain, and of 2D longitudinal strain in monoplane and triplane 2D loops were tested for agreement. For the numerical correlations of peak systolic strain (PSS) values, the Pearson correlation coefficient was used (<0.30=no correlation, 0.30 to 0.50=moderate correlation, 0.50 to 1.00=strong correlation).Results: The first numerical data evaluation in patients with normal LV wall motion was made with all PSS values, and revealed rather moderate correlations (mean Pearson coefficient 0.47, SD 0.27). Verifying both image- and tracking quality and eliminating the PSS values of segments with tracking and/or image artifacts significantly improved the numerical correlation between the methods (mean coefficient 0.68, SD 0.16). In patients with anterior myocardial infarction, pathological wall motion patterns were similarly documented by each method. The most accurate 2D-3D correlation was found between 3D area and 2D monoplane longitudinal strain (mean correlation coefficient 0.81, SD 0.12).Conclusions: Numerical strain values do not correlate perfectly between 3D and 2D strain, mainly due to issues of standardization and image quality of the 2D/3D data sets. However, the regional strain patterns between 3D area strain and 2D strain in high quality data sets show similar results in patients with normal as well as pathological regional wall motion.