A Model for Detecting Fetal Chromosomal Abnormalities Based on Quality-Weighted Regression and K-Means Clustering

Abstract

Non-invasive prenatal testing (NIPT) screens for abnormalities in chromosomes 21, 18, and 13 by analyzing fetal cell-free DNA in maternal blood. Conducted between 10 and 25 weeks of gestation, it aids in early assessment of fetal health, with accuracy contingent upon male fetuses exhibiting Y chromosome concentrations ≥4% and female fetuses displaying normal X chromosome concentrations. This paper focuses on constructing a data model for NIPT data preprocessing. A quality-weighted linear regression model was developed to establish associations between fetal Y concentration and gestational age/BMI, with statistical significance verified. Dual-breakpoint K-means clustering was used to initialize centers for BMI classification in male fetus pregnancies. Risk minimization was applied to determine the optimal NIPT timing while analyzing errors. Feature reconstruction fused LASSO regression with decision trees to establish a method for detecting female fetal chromosomal abnormalities. Final results demonstrate that this model can accurately quantify correlations between indicators, effectively capture the nonlinear relationship between gestational age and Y concentration as well as sequencing quality interference. The optimal timing for BMI groups is earlier, with high BMI groups requiring consideration of height and age. Simultaneously, it enables scientifically accurate detection of female fetal chromosomal abnormalities.