Prediction of neonatal respiratory distress syndrome by Quantitative ultrasound texture analysis of fetal lung Observational cross section study

Abstract

To predict respiratory distress syndrome using a noninvasive method called quantitative ultrasonography study of fetal lung texture. A cohort of 304 cases and was conducted in the feto maternity unit of Minia Maternity and Children University Hospital over 3 years. Fetal lung images were used to develop a computerized method based on texture analysis and machine learning algorithms, trained to predict neonatal respiratory morbidity risk on fetal lung ultrasound images. In the Control group, the range of the Replication Index (RI) was between 0.58 and 1.04, with a mean ± standard deviation (SD) of 0.82 ± 0.1. In the RD group, the RI varied from 0.53 to 1.06, with a mean ± SD of 0.77 ± 0.13. There was a statistically significant difference (p= 0.002) between the two groups. There was no notable disparity observed in the texture lung analysis when utilizing automated tools between the two groups. The research showed a significant and meaningful difference between the groups regarding Doppler indices. There was no significant differenceThere was a significant difference in occupation between the two groups tested, with a p-value of less than .001. A significant discrepancy was observed between the two analyzed groups (p= <.001). Regarding Parity, there was no statistically significant difference seen between the two groups being analyzed (p= 0.057). The research showed a significant and meaningful difference between the groups being studied in terms of LMP GA (gestational age) and AFI. Specifically, the RDS group had a much lower AFI compared to the other groups regarding texture lung analysis.