发表机构
Deutsches Herzzentrum der Charité; Charité – Universitätsmedizin Berlin; DZHK (German Centre for Cardiovascular Research); Friede Springer Cardiovascular Prevention Center; Stephenson Cardiac Imaging Centre; University of Calgary; Fraunhofer MEVIS; Physikalisch-Technische Bundesanstalt (PTB); University Hospital Tuebingen（德国心脏中心（柏林夏里特医学院）; 柏林夏里特医学院; 德国心血管研究中心; 弗里德·施普林格心血管预防中心; 斯蒂芬森心脏影像中心; 卡尔加里大学; 弗劳恩霍夫医学影像计算研究所; 德国联邦物理技术研究院; 蒂宾根大学医院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究开发了一种基于4D卷积核的全自动4D U-Net，利用稀疏标注实现4D flow MRI主动脉分割，在多中心数据上表现优异，可支持自动血流动力学分析且模型公开可用。
AI 中文摘要
4D flow MRI中的自动主动脉分割对于可重复的血流动力学评估至关重要，但受限于稀缺的密集标注和高计算需求。我们开发了一种全自动4D（3D+时间）U-Net，用于分割升主动脉、主动脉弓和近端降主动脉，采用参数高效的混合4D卷积核捕捉时间上下文，并利用从现有2D专家轮廓和中心线衍生的稀疏4D标签，从而避免了对密集4D标注的需求。训练数据包含来自8个中心、2个供应商的268次扫描，在内部测试集（32次扫描）和外部造影后测试集（30次扫描，不同站点、方案和标注者）上进行评估，与逐帧3D网络和两个半自动参考方法对比。相对于时间分辨标注，该4D U-Net的Dice分数在内部测试集为0.927，外部测试集为0.911；而3D U-Net的对应分数为0.919/0.847，静态PC-MRA为0.893，基于配准的传播方法为0.808。收缩期差异较小，舒张期差异显著。该方法与专家轮廓在峰值流速、净流量、轴向和周向壁剪切应力及直径方面的一致性极佳（内部测试集组内相关系数ICC≥0.954，外部测试集≥0.980），而半自动参考方法表现更差。因此，该方法可为自动血流动力学分析提供可重复的时间分辨主动脉分割，且可跨多中心、多供应商及独立造影后数据泛化，该模型已公开可用。
英文摘要
Automated aortic segmentation in 4D flow MRI is essential for reproducible hemodynamic assessment but is limited by scarce dense annotations and high computational demands. We developed a fully automated 4D (3D+time) U-Net for segmenting the ascending aorta, arch, and proximal descending aorta, using a parameter-efficient hybrid 4D kernel to capture temporal context and sparse 4D labels derived from existing 2D expert contours and centerlines, thereby avoiding the need for dense 4D annotations. Training comprised 268 scans from 8 centers and 2 vendors, with evaluation on an internal test set (32 scans) and an external post-contrast set (30 scans; different site, protocol, and annotator), compared against frame-wise 3D networks and two semi-automatic references. Against time-resolved annotations, the 4D U-Net achieved Dice scores of 0.927 (internal) and 0.911 (external), versus 0.919/0.847 for the 3D U-Net, 0.893 for static PC-MRA, and 0.808 for registration-based propagation; differences were small in systole but pronounced in diastole. Agreement with expert contours for peak velocity, net flow, axial and circumferential wall shear stress, and diameters was excellent (ICC >=0.954 internal, >=0.980 external), while semi-automatic references performed worse. The method thus provides reproducible, time-resolved aortic segmentation for automated hemodynamic analysis and generalizes across multicenter, multivendor, and independent post-contrast data. The model is publicly available.
CommentsSubmitted to Journal of Cardiovascular Magnetic Resonance
