DSA冠脉分割数据集

该数据集用于术中DSA影像中弹性动态血管的检测与分割性能验证。该数据集以人体冠脉血管为主要对象，涵盖该区域内复杂血管。数据内容包括2020-2021年间采集自上海华东医院的二维DSA影像，以及来自医师人工勾画的冠脉血管真实标签。标注的DSA影像数据共707帧，分辨率均为512×512，左冠和右冠的比例接近1：1。数据格式为.nii及.nii.gz，数据量为532MB。

This dataset is designed for validating the performance of elastic dynamic vascular detection and segmentation in intraoperative Digital Subtraction Angiography (DSA) images. Taking human coronary arteries as the primary target, the dataset covers complex vascular structures within this anatomical region. The data includes 2D DSA images collected at Huadong Hospital, Shanghai between 2020 and 2021, as well as the ground-truth coronary vessel annotations manually delineated by physicians. There are 707 annotated DSA image frames in total, all with a resolution of 512×512, and the ratio of left coronary artery to right coronary artery is approximately 1:1. The data is stored in .nii and .nii.gz formats, with a total size of 532 MB.

搜集汇总

背景与挑战

背景概述

该数据集是一个专注于冠脉血管分割的医学影像数据集，主要用于验证术中DSA影像中弹性动态血管的检测与分割性能。数据集包含2020-2021年从上海华东医院采集的707帧二维DSA影像，分辨率均为512×512，并附带医师人工勾画的真实标签，覆盖左冠和右冠血管，比例接近1:1，数据格式为.nii及.nii.gz，总大小为532MB。