Modeling and Optimization for Optimal Timing Decision of Non-Invasive Prenatal Testing Based on Multi-Model Fusion
Highlights in Science, Engineering and Technology · 17.04.2026 · 162 · 433-444
DOI: 10.54097/er8z9n53 ↗Kurz gefasst
Fachartikel in der Zeitschrift Highlights in Science, Engineering and Technology, online erschienen am 17.04.2026. Verfasst von Haoyu Bie und 2 weiteren Autor:innen. Die Arbeit nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
The accuracy of non-invasive prenatal testing (NIPT) is significantly affected by individual differences among pregnant women (especially body mass index, BMI) and the gestational age at which the test is performed, making it challenging to determine an individualized optimal testing time in clinical practice. To address this issue, this study aims to develop a multi-model integrated decision framework to provide precise optimal NIPT timing recommendations for pregnant women with different characteristics. Firstly, this paper employs linear mixed-effects models and gradient boosting decision trees (GBDT) to reveal the intrinsic associations between key factors—such as gestational age at testing and BMI—and fetal Y chromosome concentration from both linear and nonlinear perspectives. Based on this, for various BMI groups, a novel multi-objective optimization model integrating genetic algorithms (GA) and particle swarm optimization (PSO) is constructed, and—combined with LASSO-Cox regression for multifactorial survival analysis—achieves a transition from “prediction” to “decision-making”. Key results show that the optimized optimal testing time plan enables detection accuracy to exceed 96% for all groups of pregnant women, with a significant reduction in high-risk probability, which dropped to 0% in some groups. For female fetal samples without a Y chromosome, this study constructs a “random forest-LSTM” dual-model collaborative determination mechanism. After addressing the issue of sample imbalance using the SMOTEENN method, the model achieved a recall rate of 1.0000 and a precision rate of 0.9939 on the test set, with a missed diagnosis rate of 0%. The multi-model decision system proposed in this study provides critical theoretical and methodological support for the clinical precision application of NIPT technology and the construction of a quality control system.
Zitieren
APA
Bie, H., Ma, X., & Guo, Z. (2026). Modeling and Optimization for Optimal Timing Decision of Non-Invasive Prenatal Testing Based on Multi-Model Fusion. Highlights in Science, Engineering and Technology, 162, 433-444. https://doi.org/10.54097/er8z9n53
Vancouver
Bie H, Ma X, Guo Z. Modeling and Optimization for Optimal Timing Decision of Non-Invasive Prenatal Testing Based on Multi-Model Fusion. Highlights in Science, Engineering and Technology. 2026;162:433-444. doi:10.54097/er8z9n53
BibTeX
@article{Bie2026,
author = {Bie, Haoyu and Ma, Xu and Guo, Zixuan},
title = {Modeling and Optimization for Optimal Timing Decision of Non-Invasive Prenatal Testing Based on Multi-Model Fusion},
journal = {Highlights in Science, Engineering and Technology},
year = {2026},
volume = {162},
pages = {433--444},
doi = {10.54097/er8z9n53}
}
EUVIMED