Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
Job manager crashed while running this job (missing heartbeats).
Error code:   JobManagerCrashedError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

image
image
End of preview.

EyePACS β€” Diabetic Retinopathy Fundus Image Dataset

Merged Dataset Samples

Image: EYEPACS Preprocess Samples.


πŸ“˜ Overview

EyePACS (Eye Picture Archive Communication System) is a large-scale collection of retinal fundus images used for automated diabetic retinopathy (DR) detection.
It formed the basis of the Kaggle Diabetic Retinopathy Detection challenge, enabling research into DR classification and screening.

The dataset includes macula-centered color fundus images from real-world clinical screenings under varied imaging conditions.


πŸ“Š Dataset Summary

Field Details
Task Diabetic retinopathy classification (5 severity grades)
Description High-resolution color fundus photographs from EyePACS DR screening program. Each image labeled by ophthalmologists using the ICDR scale.
Size ~88,702 images total (β‰ˆ35k labeled for training, β‰ˆ53k unlabeled for testing)
Classes 0 = No DR, 1 = Mild, 2 = Moderate, 3 = Severe, 4 = Proliferative DR
Image Type Macula-centered color fundus photos (640Γ—480 to 5184Γ—3456 px)
Source EyePACS (USA) via Kaggle Diabetic Retinopathy Detection Challenge (2015)
Access Kaggle Dataset
License Usage restricted under Kaggle & EyePACS terms

🧱 Dataset Structure

eyepacs/
β”œβ”€β”€ images/
β”‚ β”œβ”€β”€ train/
β”‚ β”‚ β”œβ”€β”€ 00001_left.jpg
β”‚ β”‚ β”œβ”€β”€ 00001_right.jpg
β”‚ └── test/
β”œβ”€β”€ labels.csv # image_id, eye(L/R), dr_grade (0–4)
β”œβ”€β”€ README.md
└── LICENSE.txt

🧩 Label Details

  • Labels follow the International Clinical Diabetic Retinopathy (ICDR) grading system.
  • Class imbalance is significant β€” most images show no DR.
  • Labels were assigned by certified ophthalmologists; minor label noise may exist.

βš™οΈ Preprocessing Recommendations

  • Crop to the circular fundus region and remove borders
  • Resize to consistent resolution (e.g. 1024Γ—1024)
  • Normalize illumination and contrast
  • Exclude blurred or ungradable images

πŸ’‘ Research Applications

  • DR detection and severity classification
  • Automated retinal screening systems
  • Transfer learning and robustness testing across imaging conditions
  • Comparative studies with datasets like MESSIDOR, DDR, and APTOS

⚠️ Notes & Limitations

  • Significant class imbalance (majority = No DR)
  • Variations in camera type, exposure, and focus
  • Some label noise and ungradable images present
  • Redistribution may be restricted β€” verify Kaggle/EyePACS terms before publishing images

πŸ“„ Citation

If you use the dataset, cite:

Kaggle and EyePACS. β€œDiabetic Retinopathy Detection.” Kaggle Competition, 2015.
https://www.kaggle.com/c/diabetic-retinopathy-detection


πŸ”— References


Downloads last month
69

Collection including ctmedtech/EYEPACS