糖尿病视网膜病变分析挑战赛(DRAC)提供了一个基于眼底图像病灶精确分割的定量分析数据集。该数据集包含174张眼底图像,涵盖三类糖尿病视网膜病变病灶:1.视网膜内微血管异常(Intraretinal Microvascular Abnormalities);2.无灌注区(Nonperfusion Areas);3.新生血管(Neovascularization)。数据集划分为109张图像的训练集和65张图像的测试集。
yanjing维度 | 2D |
模态 | fundus |
任务类型 | segmentation |
解剖结构 | 视网膜病变 |
解剖区域 | 眼底 |
类别数 | 1 |
数据量 | 174 |
文件格式 | PNG |
DRAC2022
└── ori_dataset
└── A. Segmentation
├── 1. Original Images
│ ├── a. Training Set
│ │ ├── 065.png
│ │ ├── 074.png
│ │ └── ...
│ └── b. Testing Set
│ ├── 086.png
│ ├── 285.png
│ └── ...
└── 2. Groundtruths
└── a. Training Set
├── 1. Intraretinal Microvascular Abnormalities
│ ├── 075.png
│ ├── 082.png
│ └── ...
├── 2. Nonperfusion Areas
│ ├── 065.png
│ ├── 074.png
│ └── ...
└── 3. Neovascularization
├── 082.png
├── 083.png
└── ...
统计类型 | 间距 (mm) | 尺寸 |
---|---|---|
最小值 | - |
1024x1024x3 |
中位值 | - |
1024x1024x3 |
最大值 | - |
1024x1024x3 |
@article{qian2023drac,
title={Drac: diabetic retinopathy analysis challenge with ultra-wide optical coherence tomography angiography images},
author={Qian, Bo and Chen, Hao and Wang, Xiangning and Che, Haoxuan and Kwon, Gitaek and Kim, Jaeyoung and Choi, Sungjin and Shin, Seoyoung and Krause, Felix and Unterdechler, Markus and others},
journal={arXiv preprint arXiv:2304.02389},
year={2023}
}