A Coarse-Fine Collaborative Learning Model for Three Vessel Segmentation in Fetal Cardiac Ultrasound Images
Insights
This study introduces CoFi-Net, a deep learning tool for automatically segmenting fetal heart vessels in ultrasound images. This innovation aims to improve the accuracy and efficiency of diagnosing congenital heart disease (CHD).
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Cardiology
Background:
- Congenital heart disease (CHD) is a common birth defect and major cause of infant mortality.
- Prenatal ultrasound screening, particularly the three-vessel view (3VV), is crucial for early CHD diagnosis.
- Subjectivity and operator dependence in interpreting 3VV images limit diagnostic accuracy, especially in underserved areas.
Purpose of the Study:
- To develop an automated method for segmenting pulmonary artery, ascending aorta, and superior vena cava in the 3VV using deep learning.
- To introduce a novel deep learning network, CoFi-Net, designed for precise fetal cardiac vessel segmentation.
Main Methods:
- Proposed CoFi-Net, a deep learning network with a coarse-fine collaborative strategy.
- The network features two parallel branches: a coarse branch for global localization and a fine branch for detailed segmentation.
- Attention-parameterized skip connections were used in the fine branch to enhance feature representation and boundary details.
Main Results:
- CoFi-Net achieved superior performance in 3VV segmentation compared to existing state-of-the-art models.
- The method demonstrated significant potential for improving CHD diagnostic efficiency in clinical settings.
- CoFi-Net also showed robustness and potential for other segmentation tasks, outperforming models on a breast ultrasound dataset.
Conclusions:
- CoFi-Net offers an automated and accurate solution for 3VV segmentation in prenatal ultrasound.
- The developed deep learning approach can enhance the early diagnosis of congenital heart disease.
- CoFi-Net's versatility suggests broader applications in medical image segmentation.
Abstract:
Congenital heart disease (CHD) is the most frequent birth defect and a leading cause of infant mortality, emphasizing the crucial need for its early diagnosis. Ultrasound is the primary imaging modality for prenatal CHD screening. As a complement to the four-chamber view, the three-vessel view (3VV) plays a vital role in detecting anomalies in the great vessels. However, the interpretation of fetal cardiac ultrasound images is subjective and relies heavily on operator experience, leading to variability in CHD detection rates, particularly in resource-constrained regions. In this study, we propose an automated method for segmenting the pulmonary artery, ascending aorta, and superior vena cava in the 3VV using a novel deep learning network named CoFi-Net. Our network incorporates a coarse-fine collaborative strategy with two parallel branches dedicated to simultaneous global localization and fine segmentation of the vessels. The coarse branch employs a partial decoder to leverage high-level semantic features, enabling global localization of objects and suppression of irrelevant structures. The fine branch utilizes attention-parameterized skip connections to improve feature representations and improve boundary information. The outputs of the two branches are fused to generate accurate vessel segmentations. Extensive experiments conducted on a collected dataset demonstrate the superiority of CoFi-Net compared to state-of-the-art segmentation models for 3VV segmentation, indicating its great potential for enhancing CHD diagnostic efficiency in clinical practice. Furthermore, CoFi-Net outperforms other deep learning models in breast lesion segmentation on a public breast ultrasound dataset, despite not being specifically designed for this task, demonstrating its potential and robustness for various segmentation tasks.
