关键词: 胎儿染色体数目异常, 超声, 人工智能, 产前筛查

Abstract: Fetal chromosomal numerical abnormalities is a significant cause of pregnancy loss and birth defects. Ultrasound has emerged as a critical modality for fetal chromosomal anomaly screening due to its real-time capability, repeatability, and safety. However, its clinical application remains constrained by operator expertise variability and inconsistent image quality. The integration of artificial intelligence (AI) into conventional ultrasound has enabled the development of AI-based predictive models that overcome traditional screening limitations. These models demonstrate superior predictive performance compared to conventional methods while enabling simultaneous detection of rare chromosomal abnormalities. This review summarizes recent advances in synergistic applications of ultrasound and AI for fetal chromosomal aneuploidy prediction, comparatively analyzes the technical strengths and limitations of traditional versus AI-based predictive models, and discusses challenges including multicenter data standardization and model interpretability. These advancements provide novel directions for non-invasive precision prenatal screening.

Key words: fetal chromosomal numerical abnormalities, ultrasonography imaging, artificial intelligence, prenatal screening