Machine Learning-Based Early Prediction of Selective Fetal Growth Restriction in Monochorionic Diamniotic Twins Using Crown–Rump Length Discordance and Abnormal Placental Cord Insertion
Clinical and Experimental Obstetrics & Gynecology · 31.07.2026 · 53 (8)
DOI: 10.31083/ceog48538 ↗Kurz gefasst
Fachartikel in der Zeitschrift Clinical and Experimental Obstetrics & Gynecology, online erschienen am 31.07.2026. Verfasst von Ying Feng und 2 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Jiaxing Science and Technology Program.
Abstract (Originalsprache)
Background:To evaluate the prognostic value of crown–rump length (CRL) discordance and placental abnormal cord insertion (ACI) for predicting selective fetal growth restriction (sFGR) in monochorionic diamniotic (MCDA) twin pregnancies, and to assess their predictive performance using machine learning (ML) models.Methods:A total of 144 MCDA twin pregnancies (January 2015–December 2024) were retrospectively analyzed; 44 (30.56%) developed sFGR. According to the first-trimester ultrasound, 21 cases showed CRL discordance ≥0.60 mm, and 31 cases exhibited placental ACI. 8 supervised ML algorithms were developed and evaluated using a training set (70.14%, n = 101) and an independent validation set (29.86%, n = 43).Results:Placental ACI was significantly associated with sFGR (p < 0.01), whereas CRL discordance showed a nonsignificant positive trend. In the validation set, CRL discordance demonstrated modest discriminatory ability (area under the curve [AUC] = 0.687, 95% confidence interval [CI]: 0.4849–0.8881), whereas placental ACI showed limited discrimination based on AUC performance. For the categorical predictor (placental ACI), performance was primarily characterized using sensitivity and specificity derived from the validation set.Conclusions:CRL discordance and placental ACI provide complementary first-trimester information for early sFGR risk assessment in MCDA twins; however, overall discrimination was moderate, and external validation is warranted.
Förderung
- Jiaxing Science and Technology Program
Zitieren
APA
Feng, Y., Luo, J., & Ai, L. (2026). Machine Learning-Based Early Prediction of Selective Fetal Growth Restriction in Monochorionic Diamniotic Twins Using Crown–Rump Length Discordance and Abnormal Placental Cord Insertion. Clinical and Experimental Obstetrics & Gynecology, 53(8). https://doi.org/10.31083/ceog48538
Vancouver
Feng Y, Luo J, Ai L. Machine Learning-Based Early Prediction of Selective Fetal Growth Restriction in Monochorionic Diamniotic Twins Using Crown–Rump Length Discordance and Abnormal Placental Cord Insertion. Clinical and Experimental Obstetrics & Gynecology. 2026;53(8). doi:10.31083/ceog48538
BibTeX
@article{Feng2026,
author = {Feng, Ying and Luo, Jing and Ai, Ling},
title = {Machine Learning-Based Early Prediction of Selective Fetal Growth Restriction in Monochorionic Diamniotic Twins Using Crown–Rump Length Discordance and Abnormal Placental Cord Insertion},
journal = {Clinical and Experimental Obstetrics & Gynecology},
year = {2026},
volume = {53},
number = {8},
doi = {10.31083/ceog48538}
}
EUVIMED