RETA Dataset
Images
RETA contains 81 images inherited from the first subset of IDRiD dataset. 54 images for training and 27 images for testing.
Labels
- Pixel-level blood vessel masks (Current available labels. Note: testing labels are for online evaluation only.).
- Artery/vein masks, vascular skeletons, bifurcations, trees and abnormalities.
Annotation Quality Evaluation
See our publication for details.
Download
RETA dataset is available at https://doi.org/10.6084/m9.figshare.16960855
This work is licensed under a Creative Commons Attribution 4.0 International License.
Citation
Lyu, X., Cheng, L. & Zhang, S. The RETA Benchmark for Retinal Vascular Tree Analysis.
Sci Data 9, 397 (2022). https://doi.org/10.1038/s41597-022-01507-y
Other Public Datasets
Retinal vessel segmentation/classification in color fundus images:
- DRIVE, http://www.isi.uu.nl/Research/Databases/DRIVE/
- STARE, http://cecas.clemson.edu/~ahoover/stare/
- ARIA, http://www.damianjjfarnell.com/?page_id=276
- CHASE_DB1, https://blogs.kingston.ac.uk/retinal/chasedb1/
- HRF, https://www5.cs.fau.de/research/data/fundus-images/
- DualModal2019, https://ieee-dataport.org/documents/dualmodal2019-dataset
- DRHAGIS, http://personalpages.manchester.ac.uk/staff/niall.p.mcloughlin/
- UoA-DR, https://auckland.figshare.com/articles/UoA-DR_Database_Info/5985208/5
- LES-AV, https://github.com/ignaciorlando/glaucoma-hemodynamics
- ORVS, https://github.com/AbdullahSarhan/ICPRVessels
- AFIO, https://data.mendeley.com/datasets/3csr652p9y/1
Latest datasets:
- TREND, https://zenodo.org/record/4521044
- FIVES, https://figshare.com/articles/figure/FIVES_A_Fundus_Image_Dataset_for_AI-based_Vessel_Segmentation/19688169
Besides, retinal vessel segmentation/classification in other imaging modalities include