Role of artificial-intelligence-assisted automated cardiac biometrics in prenatal screening for coarctation of aorta

C. A. Taksøe-Vester*, K. Mikolaj, O. B.B. Petersen, N. G. Vejlstrup, A. N. Christensen, A. Feragen, M. Nielsen, M. B.S. Svendsen, M. G. Tolsgaard

*Corresponding author af dette arbejde

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

1 Citationer (Scopus)
1 Downloads (Pure)

Abstract

Objective
Although remarkable strides have been made in fetal medicine and the prenatal diagnosis of congenital heart disease, around 60% of newborns with isolated coarctation of the aorta (CoA) are not identified prior to birth. The prenatal detection of CoA has been shown to have a notable impact on survival rates of affected infants. To this end, implementation of artificial intelligence (AI) in fetal ultrasound may represent a groundbreaking advance. We aimed to investigate whether the use of automated cardiac biometric measurements with AI during the 18–22-week anomaly scan would enhance the identification of fetuses that are at risk of developing CoA.

Methods
We developed an AI model capable of identifying standard cardiac planes and conducting automated cardiac biometric measurements. Our data consisted of pregnancy ultrasound image and outcome data spanning from 2008 to 2018 and collected from four distinct regions in Denmark. Cases with a postnatal diagnosis of CoA were paired with healthy controls in a ratio of 1:100 and matched for gestational age within 2 days. Cardiac biometrics obtained from the four-chamber and three-vessel views were included in a logistic regression-based prediction model. To assess its predictive capabilities, we assessed sensitivity and specificity on receiver-operating-characteristics (ROC) curves.

Results
At the 18–22-week scan, the right ventricle (RV) area and length, left ventricle (LV) diameter and the ratios of RV/LV areas and main pulmonary artery/ascending aorta diameters showed significant differences, with Z-scores above 0.7, when comparing subjects with a postnatal diagnosis of CoA (n = 73) and healthy controls (n = 7300). Using logistic regression and backward feature selection, our prediction model had an area under the ROC curve of 0.96 and a specificity of 88.9% at a sensitivity of 90.4%.

Conclusions
The integration of AI technology with automated cardiac biometric measurements obtained during the 18–22-week anomaly scan has the potential to enhance substantially the performance of screening for fetal CoA and subsequently the detection rate of CoA. Future research should clarify how AI technology can be used to aid in the screening and detection of congenital heart anomalies to improve neonatal outcomes. © 2024 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.
OriginalsprogEngelsk
TidsskriftUltrasound in Obstetrics and Gynecology
Vol/bind64
Udgave nummer1
Sider (fra-til)36-43
ISSN0960-7692
DOI
StatusUdgivet - 2024

Bibliografisk note

Funding Information:
The Project is supported by The Novo Nordisk Foundation through grant NNFSA170030576, the Danish Regions' AI Signature Project, through the Centre for Basic Machine Learning Research in Life Science (NNF20OC0062606) and the Pioneer Centre for AI, DNRF grant no. P1.

Publisher Copyright:
© 2024 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

Citationsformater