Predictors of Missing Outputs in AI-Based Fetal CNS Biometry Using SonoCNS: A Real-World Cohort Study
Journal of Clinical Medicine · 15.11.2026 · 15 (22) · 8946
DOI: 10.3390/jcm15228946 ↗Kurz gefasst
Kohortenstudie in der Zeitschrift Journal of Clinical Medicine, online erschienen am 15.11.2026. Verfasst von Jakub Mlodawski und 4 weiteren Autor:innen. Die Arbeit nennt 17 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Creators Fund of the KOPIPOL Association.
Abstract (Originalsprache)
Background
Artificial intelligence (AI)-based automated fetal central nervous system (CNS) biometry using SonoCNS may streamline ultrasound assessment, but missing outputs occur in clinical practice. This study aimed to quantify the frequency and patterns of missing SonoCNS measurements across two separately acquired fetal head volumes and to investigate their associations with gestational age, operator, fetal presentation, and fetal biometric characteristics.
Methods
This single-center retrospective secondary analysis used prospectively collected data from 385 women examined between 1 May 2022 and 31 May 2023. Each participant underwent second- or third-trimester ultrasound assessment with SonoCNS measurements obtained from two separately acquired volumes (a1, a2) for BPD, HC, OFD, cisterna magna (CM), lateral ventricular atrial width (Vp), and transverse cerebellar diameter (Cereb). Primary endpoints (full set: BPD/HC/OFD/CM/Vp/Cereb) were persistent AI failure (the same parameter missing in both acquisitions for at least one of the six parameters) and AI instability (missing ≥1 parameter in ≥1 acquisition). Structure-specific models assessed miss_CM, miss_Vp, and miss_Cereb. Logistic regression adjusted for operator, fetal presentation, gravidity/parity, and GA-adjusted fetal size residual z-scores (EFW-, HC-, and BPD-for-GA); sensitivity analyses were restricted to AZZ and to examinations at
Förderung
- Creators Fund of the KOPIPOL Association
Zitieren
APA
Mlodawski, J., Zmelonek-Znamirowska, A., Detka, K., Mlodawska, M., & Swiercz, G. (2026). Predictors of Missing Outputs in AI-Based Fetal CNS Biometry Using SonoCNS: A Real-World Cohort Study. Journal of Clinical Medicine, 15(22), 8946. https://doi.org/10.3390/jcm15228946
Vancouver
Mlodawski J, Zmelonek-Znamirowska A, Detka K, Mlodawska M, Swiercz G. Predictors of Missing Outputs in AI-Based Fetal CNS Biometry Using SonoCNS: A Real-World Cohort Study. Journal of Clinical Medicine. 2026;15(22):8946. doi:10.3390/jcm15228946
BibTeX
@article{Mlodawski2026,
author = {Mlodawski, Jakub and Zmelonek-Znamirowska, Anna and Detka, Karolina and Mlodawska, Marta and Swiercz, Grzegorz},
title = {Predictors of Missing Outputs in AI-Based Fetal CNS Biometry Using SonoCNS: A Real-World Cohort Study},
journal = {Journal of Clinical Medicine},
year = {2026},
volume = {15},
number = {22},
pages = {8946},
doi = {10.3390/jcm15228946}
}
EUVIMED