Related Experiment Video
Updated: Sep 9, 2025
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
Published on: July 5, 2024
A multi-backbone cascade and morphology-aware segmentation network for complex morphological X-ray coronary artery
Xiaodong Zhou1, Huibin Wang2, Lili Zhang2
1College of Artificial Intelligence and Automation, Hohai University, Nanjing, 210000, Jiangsu, China.
Insights
This study introduces MBCMA-Net, a novel deep learning model for segmenting coronary arteries in X-ray images. The network achieves superior accuracy and connectivity, improving diagnosis of coronary artery disease.
Area of Science:
- Medical Imaging
- Computer Vision
- Artificial Intelligence
Background:
- X-ray coronary angiography is the gold standard for diagnosing coronary artery disease.
- Existing segmentation methods struggle with complex coronary artery morphology, leading to errors and vessel breakage.
Purpose of the Study:
- To develop an advanced segmentation network, MBCMA-Net, for accurate coronary artery analysis.
- To improve feature extraction and complex structure recognition in coronary angiography.
Main Methods:
- Proposed a multi-backbone cascade and morphology-aware segmentation network (MBCMA-Net).
- Incorporated multi-backbone encoders for enhanced feature extraction.
- Integrated a vascular morphology-aware module for complex structure recognition.
- Utilized a centerline loss function to maintain vascular connectivity.
Main Results:
- MBCMA-Net achieved an IoU of 87.14% and a DSC score of 92.72%.
- Demonstrated superior vascular connectivity with a score of 89.05%.
- Outperformed mainstream segmentation algorithms on public datasets (DCA1, CADICA, ARCADE).
Conclusions:
- MBCMA-Net offers a robust solution for coronary artery segmentation.
- The model can serve as a benchmark for future coronary artery segmentation research.
- Improved segmentation accuracy aids in better diagnosis of coronary artery disease.
Abstract:
X-ray coronary artery images are the 'gold standard' technology for diagnosing coronary artery disease, but due to the complex morphology of the coronary arteries, such as overlapping, winding and uneven contrast media filling, the existing segmentation methods often suffer from segmentation errors and vessel breakage. To this end, we proposed a multi-backbone cascade and morphology-aware segmentation network (MBCMA-Net), which improves the feature extraction ability of the network through multi-backbone encoders, and embeds a vascular morphology-aware module in the backbone network to enhance the capability of complex structure recognition, and finally introduces a centerline loss function to maintain the vascular connectivity. During the experiment, we selected 1942 clear angiograms from two public datasets (DCA11 and CADICA2) and annotated them, and also used the public ARCADE3 dataset for testing. Experimental results show that MBCMA-Net reaches an IoU of 87.14%, a DSC score of 92.72%, and a vascular connectivity score of 89.05%, which is better than the mainstream segmentation algorithms and can be used as a benchmark model for coronary artery segmentation. Code repository: https://gitee.com/zaleman/mbcma-net.
