Automated Detection of Congenital Heart Disease in Fetal Ultrasound Screening

TL;DR

This paper explores deep learning methods to automate the detection of congenital heart disease in fetal ultrasound, aiming to improve screening accuracy and reduce reliance on human expertise.

Contribution

It introduces a novel pipeline with auxiliary view classification to enhance deep learning detection of CHD in fetal ultrasound images.

Findings

F1-score improved from 0.72 to 0.87 for healthy cases.

F1-score improved from 0.77 to 0.85 for CHD cases.

Demonstrates potential for AI-assisted prenatal screening.

Abstract

Prenatal screening with ultrasound can lower neonatal mortality significantly for selected cardiac abnormalities. However, the need for human expertise, coupled with the high volume of screening cases, limits the practically achievable detection rates. In this paper we discuss the potential for deep learning techniques to aid in the detection of congenital heart disease (CHD) in fetal ultrasound. We propose a pipeline for automated data curation and classification. During both training and inference, we exploit an auxiliary view classification task to bias features toward relevant cardiac structures. This bias helps to improve in F1-scores from 0.72 and 0.77 to 0.87 and 0.85 for healthy and CHD classes respectively.