Explore high-quality datasets for your AI and machine learning projects.
TJDR is a high‑quality diabetic retinopathy pixel‑level annotation dataset comprising 561 color fundus images from Tongji Hospital, Tongji University. Images were captured with various fundus cameras at high resolution. Personal identifiers have been removed. Anatomical structures such as optic disc, retinal vessels, and macula are clearly visible. Experienced ophthalmologists annotated four common DR lesion types—micro‑aneurysms (MA), hemorrhage (HE), hard exudates (EX), and soft exudates (SE)—using the Labelme tool. The dataset is split into training and test sets and publicly released to support DR lesion segmentation research.
