Novel Speckle Tracking Analysis Showed Excellent Reproducibility for Size and Shape of the Fetal Heart and Good Reproducibility for Strain and Fractional Shortening

Abstract

Introduction: This study aimed to evaluate reproducibility and agreement of fetal cardiac shape and deformation using FetalHQ. Methods: Fifty normal fetuses at 20–38 weeks of gestation were evaluated. Two operators independently selected an optimal cardiac cycle using FetalHQ^®™software for speckle tracking analysis. Intra- and interobserver correlation coefficient and limits of agreement for cardiac shape and deformation were estimated. Results: Global cardiac markers: high correlation (r = 0.98) and agreement (mean difference, standard deviation [MD, SD] 5.07, 75.8) for ventricular area; moderate correlation (r = 0.78) and agreement (MD, SD: 0.016, 0.08) for global sphericity index (SI) and for left ventricle (LV) global strain (r = 0.65; MD, SD: −4.48, 11.9); and low but still significant correlation (r = 0.58) and agreement (MD, SD: −3.77, 12.27) for right ventricle (RV) global strain. For individual ventricular parameters: high correlation for LV ([median r; range] 0.98; 0.93–0.99) and RV (r = 0.98; 0.97–1.0) SI, and for LV (r = 0.92: 0.56–0.99) and RV (r = 0.96; 0.67–0.99) end diastolic diameters; moderate correlation for LV fractional shortening (r = 0.53; 0.87–0.98); and no significant correlation for RV fractional shortening (r = 0.36; 0.32–0.97). Inter- and intraobserver correlation and agreement were similar for all evaluated parameters. Conclusion: Speckle tracking analysis of the fetal heart provides reliable estimations of global and LV shape and deformation. Low correlation in the RV can be related to anatomical structures such as the moderator band.