Reproducibility and Optimization of Analysis Parameters of Tissue Doppler-Derived Strain and Strain Rate Measurements for Very Low Birth Weight Infants

Abstract

To assess the effects of region of interest (ROI) size and strain length (SL) on the quality of analysis and to determine the feasibility and reproducibility of tissue Doppler-derived cardiac strain (S) and strain rate (SR) measurements for preterm infants <1500g birth weight. The beat-to-beat variation (BBV) for different combinations of ROI width (RW), ROI length (RL), and SL was determined from 60 good quality images of left and right free walls and the interventricular septum. We examined the impact of RW, RL, and SL on BBV and estimated inter- and intra-operator variability. BBV was smallest for RW=2mm. Raising SL (even at the cost of RL) led to a decrease in BBV. Several combinations provided a low BBV without significant difference. Coefficients of variation for intra- and inter-operator variability ranged from 15.1% to 65.2%. Both were lower for systolic than for diastolic values. Tissue Doppler imaging-derived S and SR measurements are feasible in very low birth weight infants. Reproducibility is comparable to older infants for systolic S and SR. For optimal analysis quality, SL and RW should be high while RL should be small (though inside segment size). Best results were obtained with RW=2mm, RL=1mm, and SL=6mm.