Feasibility and Reproducibility of a Structure-Guided Deep Learning Model for Automatic Detection of the Standard Sagittal Plane in First-Trimester Nuchal Translucency Assessment Using 3D Ultrasound
An automated AI system measures first-trimester neck thickness from 3D ultrasound as reliably as expert manual methods, enabling standardized screening.
A prospective study developed and validated 3D MSP-net, a CNN model using intracranial landmark segmentation for automated mid-sagittal plane extraction from 3D NT ultrasound volumes, achieving 91.6% success comparable to expert manual measurement and outperforming commercial rule-based alternatives. NT measurements were equivalent to the conventional manual method, supporting potential for standardization of first-trimester fetal screening.
What the study was
- Study design
- Prospective diagnostic model development and validation study
- Population
- Singleton pregnant women undergoing first-trimester NT screening (Yonsei University Health System)
- Category
- Diagnostics
- Maturity
- Exploratory
- Journal
- Journal of Ultrasound Medicine
Why it surfaced
Prospective AI model for first-trimester fetal screening standardization — addresses a persistent clinical challenge in prenatal ultrasound quality. Promising but sample size not reported; single-center; external validation needed.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.