Advanced Segmentation of X-Ray Coronary Angiography: Leveraging DA-TransUNet for Enhanced Vessel Visualization

We used Automatic Region-based coronary artery disease diagnostics using X-Ray angiography imagEs (ARCADE) dataset, publicly available . In this work we applied ARCADE dataset that includes 3,000 XCA images were equally divided for artery classification and stenosis detection, with 1,500 images allocated to each category. The ARCADE dataset was used to diagnose CAD, emphasizing coronary channel tree segmentation with the SYNTAX Score and stenosis detection. For both tasks 1,000 images allocated for training, 200 allotted for validation, and 300 reserved for test.

我们采用了基于自动区域分析的X线冠状动脉造影疾病诊断（Automatic Region-based coronary artery disease diagnostics using X-Ray angiography imagEs，简称ARCADE）公开数据集。本研究使用该ARCADE数据集，其包含3000幅X线冠状动脉造影（X-ray Coronary Angiography，简称XCA）图像，按任务类型均等划分为动脉分类与狭窄检测两个子集，每个子集各含1500幅图像。该数据集被用于冠状动脉疾病（Coronary Artery Disease，简称CAD）的诊断，研究重点围绕冠状动脉树分割、SYNTAX评分计算及狭窄检测展开。针对上述两项任务，数据集均按照1000幅用于训练、200幅用于验证、300幅用于测试的比例进行划分。