Feasibility and Reproducibility of a Standard Protocol for 2D Speckle Tracking and Tissue Doppler-based Strain and Strain Rate Analysis of the Fetal Heart

Purpose: Assessment of cardiac function in the fetal heart is challenging because of its small size and high heart rate, restricted physical access to the fetus, and impossibility of fetal ECG recording. We aimed to standardize the acquisition and postprocessing of fetal echocardiography for deformation analysis and to assess its feasibility, reproducibility, and correlation for longitudinal strain and strain rate measurements by tissue Doppler imaging (TDI) and 2D speckle tracking (2D-strain) during pregnancy.

Methods: Echocardiography was performed in 56 fetuses. 2D and color TDI in apical or basal four-chamber views were recorded for subsequent analysis. Caution was taken to achieve a frame rate >70 Hz for speckle tracking and >150 Hz for TDI analysis. For each acquisition, 7.5 s of noncompressed data were stored in cine loop format and analyzed offline. Since fetal ECG information is by definition not available, aortic valve closure was marked from aortic flow and the onset of each cardiac cycle was manually indicated in the 2D images. Sample volume length was standardized at the minimum size. Two observers measured the left and right ventricular peak systolic longitudinal strain and strain-rate.

Results: Strain and strain rate measurements were feasible in 93% of the TDI and 2D-strain acquisitions. The mean time spent on analyzing TDI images was 18 min, with an intraclass agreement coefficient of 0.86 (95% CI 0.77-0.92), 0.83 (95% CI 0.72-0.90), 0.96 (95% CI 0.93-0.98), and 0.86 (95% CI 0.76-0.92) for basal left and right free wall peak systolic strain and strain rate, respectively. Agreement between observers using tissue Doppler also showed high reliability. The mean time spent for 2D-strain analysis was 15 min, with an intraclass agreement coefficient of 0.97 (95% CI 0.95-0.98), 0.94 (95% CI 0.89-0.96), 0.96 (95% CI 0.93-0.98), and 0.84 (95% CI 0.73-0.90) for basal left and right free wall peak systolic strain and strain rate, respectively. Agreement between observers also showed a high reliability that was similar for TDI and 2D-strain. There was a weak correlation between TDI and 2D-strain measurements.

Conclusions: A standard protocol with fixed acquisition and processing settings, including manual indication of the timing events of the cardiac cycle to correct for the lack of ECG, was feasible and reproducible for the evaluation of longitudinal ventricular strain and strain rate of the fetal heart by TDI as well as 2D-strain analysis. However, both techniques are not interchangeable as the correlation between them is relatively poor.