Abstract
PURPOSE: To evaluate differences in strain measurements by using tissue-tracking (TT) and feature-tracking (FT) software on cardiovascular MR images in patients with repaired tetralogy of Fallot (TOF). MATERIALS AND METHODS: In this retrospective cross-sectional study of 25 patients with repaired TOF (median age, 33.1 years; interquartile range, 25.3-38.3 years) from 2008 through 2014, left ventricular (LV) and right ventricular (RV) global circumferential and longitudinal strain were measured from cardiac MR images by using TT and FT software. Time to process was measured from opening the study to acceptance of contours. Intra- and interobserver reproducibility were evaluated with Bland-Altman analysis, coefficient of variation, and intraclass correlation coefficient. RESULTS: Time to process was slightly longer for TT (10.2 minutes ± 3.1 [standard deviation] vs 9.0 minutes ± 1.7, P = .04). Fewer patients required contour revision with TT than with FT. Both TT and FT measurements had similar moderate-to-strong correlations with LV and RV ejection fractions; correlation of RV longitudinal strain with RV ejection fraction did not reach significance by using either method. With the exception of LV circumferential strain, strain measurements were lower with FT relative to TT. Intra- and interobserver reproducibility were lower with FT for longitudinal strain measurements. CONCLUSION: TT and FT have systematic differences in strain values and reproducibility, particularly for longitudinal strain. Software-specific normative data are necessary, as are studies to evaluate correlation with clinical outcomes for each modality.© RSNA, 2019.