Development of Digital Fetal Heart Models with Virtual Ultrasound Function Based on Cardiovascular Casting and Computed Tomography Scan

Abstract

Despite recent breakthroughs in diagnosis and treatment, congenital heart defects (CHDs) continue to be the leading cause of death among newborns. Fetal echocardiography is the most effective and non-invasive method for the prenatal diagnosis of CHDs. However, the challenge of obtaining standard views can lead to a low diagnostic accuracy. To explore new methods for training, the combined use of cardiovascular casting, computed tomography (CT) scanning, and virtual ultrasound generation methods was studied to preserve the cardiac structures of a fetus in digital form. The feasibility of the proposed workflow was verified by testing three fetal heart specimens collected after the termination of pregnancy. As a result, the anatomical structures were imaged clearly by a CT scan after cardiovascular casting, and the virtually generated ultrasound images based on the use of the Public software Library for UltraSound imaging research (PLUS) toolkit successfully demonstrated both the standard views and the views with diagnostic values for the visualization of the cardiovascular structures. This solution provides great data extensibility while being simple and cost-effective for end users. Therefore, the proposed method could provide a promising educational system for trainees to understand standard views of fetal echocardiography and the corresponding anatomical correlations.

Applications of artificial intelligence-powered prenatal diagnosis for congenital heart disease

“…Due to limitations on radiation exposure during pregnancy, invasive examinations such as CT scans cannot be performed. To achieve three-dimensional imaging of the fetal heart, Yang et al ( 32 ) proposed a workflow that combines post-mortem fetal heart and cardiovascular casting with CT scanning and fetal echocardiography. The process includes obtaining standard views for alignment and then performing a three-dimensional reconstruction.…”

Improving the Quality of Fetal Heart Ultrasound Imaging With Multihead Enhanced Self-Attention and Contrastive Learning

“…Recently, convolutional neural networks (CNNs) have shown good performance in automatic analyses of fetal echocardiography data, including optimizing the scanning process, obtaining standard views, and assessing image quality. For example, Yang et al [26] used postmortem fetal heart and cardiovascular casts combined with CT scans and fetal echocardiography data to determine the pose relationship between each section and optimize the data acquisition process. Chen et al [25] explored a neural network to automatically detect standard views from fetal heart ultrasound scan videos and selected different views by using a threshold value.…”