posted on 2022-05-01, 00:57authored byKai JinKai Jin, Xingru Huang, Jingxin ZhouJingxin Zhou, Yunxiang Li, Yan Yan, Yibao Sun, Qianni Zhang, Yaqi Wang, Juan Ye
FIVES dataset consists of 800 high-resolution multi-disease color fundus photographs with pixel-wise manual annotation. The annotation process was standardized through crowdsourcing of a group of medical experts. The quality of each image was evaluated, including illumination and color distortion, blur, and low contrast distortion, based on which the data splitting was conducted to make sure the balanced distribution of image features.
Detailed descriptions can be found in the original paper, and please cite it if utilizing any part of the dataset:
Jin, K., Huang, X., Zhou, J. et al. FIVES: A Fundus Image Dataset for Artificial Intelligence based Vessel Segmentation. Sci Data9, 475 (2022). https://doi.org/10.1038/s41597-022-01564-3