SPReCHD: Four-Chamber Semantic Parsing Network for Recognizing Fetal Congenital Heart Disease in Medical Metaverse
Related Experiment Video
Updated: Aug 23, 2025
Anatomically Realistic Neonatal Heart Model for Use in Neonatal Patient Simulators
Published on: February 5, 2019
Insights
This study introduces SPReCHD, a novel network for fetal congenital heart disease (CHD) detection. SPReCHD significantly improves prenatal CHD recognition by accurately analyzing fetal heart structures from ultrasound images.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Cardiology
Background:
- Congenital heart disease (CHD) is a common congenital malformation requiring early detection.
- Fetal echocardiography is crucial for CHD screening, but faces challenges like fetal movement and image artifacts, leading to low prenatal detection rates.
- Accurate analysis of fetal cardiac structures is vital for timely diagnosis and intervention.
Purpose of the Study:
- To develop an advanced deep learning model for improved fetal CHD recognition from ultrasound images.
- To enhance the segmentation and localization of fetal four-chamber cardiac structures.
- To increase the prenatal detection rate of fetal CHD.
Main Methods:
- A multi-scale gated axial-transformer network (MSGATNet) was developed to capture semantic information from fetal four-chamber views.
- The SPReCHD network integrates MSGATNet for segmenting and locating arbitrary four-chamber contours.
- The model was trained and validated on ultrasound datasets to evaluate its performance in recognizing fetal CHD.
Main Results:
- The SPReCHD network achieved high performance metrics on the test set: 95.92% precision, 94% recall, 95% accuracy, and 94.95% F1-score.
- The proposed method effectively captures discriminative representations of the fetal heart, overcoming limitations of traditional methods.
- Significant improvement in the prenatal detection rate of fetal CHD was demonstrated.
Conclusions:
- The SPReCHD network, integrating MSGATNet, offers a robust solution for fetal CHD recognition in medical metaverse applications.
- The model's ability to accurately segment and analyze fetal cardiac structures leads to a substantial increase in prenatal CHD detection.
- This advancement holds promise for improving clinical management and outcomes for fetuses diagnosed with CHD.
Abstract:
Echocardiography is essential for evaluating cardiac anatomy and function during early recognition and screening for congenital heart disease (CHD), a widespread and complex congenital malformation. However, fetal CHD recognition still faces many difficulties due to instinctive fetal movements, artifacts in ultrasound images, and distinctive fetal cardiac structures. These factors hinder capturing robust and discriminative representations from ultrasound images, resulting in CHD's low prenatal detection rate. Hence, we propose a multi-scale gated axial-transformer network (MSGATNet) to capture fetal four-chamber semantic information. Then, we propose a SPReCHD: four-chamber semantic parsing network for recognizing fetal CHD in the clinical treatment of the medical metaverse, integrating MSGATNet to segment and locate four-chamber arbitrary contours, further capturing distinguished representations for the fetal heart. Comprehensive experiments indicate that our SPReCHD is sufficient in recognizing fetal CHD, achieving a precision of 95.92%, a recall of 94%, an accuracy of 95%, and a F1 score of 94.95% on the test set, dramatically improving the fetal CHD's prenatal detection rate.
