Quantitative Ultrasound Texture Analysis for Differentiating Preterm From Term Fetal Lungs.

摘要: Objectives To differentiate preterm (<37 weeks' gestation) from term (≥37 fetal lungs by using quantitative texture analysis of ultrasound images. Methods This study retrospectively evaluated singleton gestations with valid dating at 20 gestational age (GA) or later between January 2015 and December 2015. Images were obtained Voluson E8 systems (GE Healthcare, Milwaukee, WI). A region interest was selected in each lung image the level 4 heart chambers an area that appeared most representative overall tissue had least shadow. Ultrasonic heterogeneity (heterogeneity index) based on dynamic range calculation determined for all images. This quantification performed a custom-made software program used dithering technique Floyd-Steinberg algorithm, which pixels are transformed into binary map. Regression to determine correlation functional association index GA. receiver operating characteristic curve identify optimal cutoff point differentiating lungs. Results A total 425 images (313 112 term) analyzed. Quantitative predicted GA sensitivity specificity 87.9% 92.0%, respectively, point. Conclusions Quantitative can lungs. Our data suggest decreased imaging is associated fetuses.