Strain Analysis by Tissue Doppler Imaging: Comparison of Conventional Manual Measurement with a Semiautomated Approach

PURPOSE: Myocardial strain and strain rate (SR) can be derived from either tissue Doppler (TDI) information or two-dimensional speckle tracking. As conventional TDI analysis (TDI-manual) is time-consuming with poor reproducibility, we developed a faster semiautomated approach (TDI-ST). We aimed to study the comparability of TDI-ST with TDI-manual, an established method for measuring strain and SR. METHODS: Forty healthy subjects (mean age 38.3 ± 12.8 years) and 16 patients with FHL-1 cardiomyopathy (CMP) (36.8 ± 14.2 years) were analyzed with TDI-manual and TDI-ST. TDI-ST was performed with commercial software, using speckle tracking for myocardial tracking and TDI information to derive longitudinal strain and SR from high frame rate TDI recordings. Measurements of longitudinal systolic strain (S) and global S (GLS) made with the two methods were compared with Bland-Altman plots and Deming regression. Receiver operating characteristics (ROC) curves were used to compare discrimination between healthy individuals and patients. RESULTS: Mean S was -20.11 ± 4.85% (healthy) and -16.12 ± 4.44% (CMP) with TDI-ST and -21.15 ± 5.68% (healthy) and -16.27 ± 6.44 (CMP) with TDI-manual. Using all measured segments, the mean bias was 0.78% strain toward less negative S with TDI-ST; the Deming regression slope was 0.7 for S and 0.9 for GLS. Intra- and inter-observer CVs were 5.4% and 7.0%, respectively. ROC curves showed no significant differences between the methods. CONCLUSION: The described S and SR measurements with TDI-ST are comparable to conventional manual analysis. Thus, using TDI-ST, it is possible to quickly and easily extract high-resolution deformation data.