基于融合双重注意力机制的心脏多结构细分网络
Guodong Zhang1, Luchang Yang1, Yanlin Li1
1School of Computer Science, Shenyang Aerospace University, Shenyang, 110136, China.
Medical & biological engineering & computing
|February 9, 2026
概括
这项研究引入了一种新的3D深度学习网络,用于CT图像中准确的心脏细分,通过精确量化心脏功能指标来改善心血管疾病的诊断.
科学领域:
- 医疗成像医学成像
- 人工智能的人工智能
- 心血管疾病 心血管疾病
背景情况:
- 准确的心脏细分对于诊断心血管疾病至关重要.
- 计算机断层扫描 (CT) 图像带来了诸如模糊边界和类似组织强度等挑战,阻碍了细分.
- 现有的方法难以准确地划分心脏结构.
研究的目的:
- 开发一个先进的3D深度学习网络,用于CT图像中的多结构心脏细分.
- 通过改进的细分来提高心脏功能量化的准确性.
- 为了克服CT扫描中模糊边界和相邻组织干扰的局限性.
主要方法:
- 提出了一个3D心脏多结构细分网络,包括多尺度通道增强注意力 (MCEA) 和空间分解与通道融合注意力 (SD-CA) 模块.
- MCEA通过整合不同尺度的道信息来增强多尺度的特征表示.
- SD-CA生成并行空间和频道注意力权重,专注于感兴趣的区域并减少干扰.
主要成果:
- 实现了左心室,心肌,左心室,右心室和右心室的高细分精度,平均子系数为94.21%和93.9%.
- 展示了强大的性能,平均95%的豪斯多夫距离为6.5483和4.36.
- 左心室喷射分数 (LVEF) 和亚结构体积的定量预测显示相关系数>0.9587和布兰德-阿尔特曼测试中的高一致性 (>94.8%).
结论:
- 拟议的3D网络有效地解决了心脏CT图像细分方面的挑战,特别是模糊的边界和类似的强度.
- MCEA和SD-CA模块显著提高了细分性能和量化心脏功能指标的能力.
- 结果表明,它对心血管疾病的临床诊断和治疗计划具有很高的准确性和可靠性.
更多相关视频
相关概念视频
Mechanism of Cardiac Arrhythmias
2.3K
Arrhythmias are irregular heart rhythms occurring when the heart's electrical impulses become abnormal. These disturbances can lead to various symptoms, depending on their severity and the underlying cause. Some common factors contributing to arrhythmias include hypoxia, ischemia, electrolyte imbalances, excessive catecholamine exposure, drug toxicity, and muscle overstretching. Arrhythmias can be classified into two main types based on the rate and site of origin of abnormal heart rhythms.
2.3K
Structure of Cardiac Muscles
17.1K
Cardiac muscle, or myocardium, is a specialized type of muscle found exclusively in the heart. Its unique structural and functional characteristics enable the heart to perform its vital role of pumping blood throughout the body continuously and rhythmically. The cardiac muscle cells, or cardiomyocytes, possess an endomysium and perimysium but do not have an epimysium.
Compared to skeletal muscles, cardiac muscle cells are small and mostly have a single nucleus. Additionally, they are usually...
Compared to skeletal muscles, cardiac muscle cells are small and mostly have a single nucleus. Additionally, they are usually...
17.1K
Protein Networks
4.6K
An organism can have thousands of different proteins, and these proteins must cooperate to ensure the health of an organism. Proteins bind to other proteins and form complexes to carry out their functions. Many proteins interact with multiple other proteins creating a complex network of protein interactions.
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
4.6K
Reclosers and Fuses
485
Automatic circuit reclosers enhance the protection of distribution circuits by interrupting and auto-reclosing an AC circuit according to a preset sequence. They effectively manage temporary faults on overhead distribution lines, often caused by tree limbs or wildlife, by briefly disrupting service to improve overall reliability. However, contact with reclosers or energized broken conductors on the ground can pose serious hazards.
A comprehensive protection scheme for radial distribution...
A comprehensive protection scheme for radial distribution...
485
Circuit Breaker and Fuse Selection
616
A circuit breaker is a device engineered to interrupt fault currents and sometimes reclose automatically. When a fault current is detected, the breaker separates the electrical contacts, which generates an arc. This arc is extinguished by methods such as elongation, cooling, or splitting, depending on the breaker's design. Breakers are categorized based on the voltage they operate at and the medium used for arc extinction, such as air, oil, SF6 gas, or vacuum.
In high-voltage systems,...
In high-voltage systems,...
616
Network Covalent Solids
16.2K
Network covalent solids contain a three-dimensional network of covalently bonded atoms as found in the crystal structures of nonmetals like diamond, graphite, silicon, and some covalent compounds, such as silicon dioxide (sand) and silicon carbide (carborundum, the abrasive on sandpaper). Many minerals have networks of covalent bonds.
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
16.2K


