Published June 4, 2026
| Version v3
Dataset
Open
CornOrb: A Multimodal Dataset of Orbscan Corneal Topography and Clinical Annotations for Keratoconus Detection
Authors/Creators
-
1.
University of Abou Bekr Belkaïd
-
2.
Université de Bretagne Occidentale
Contributors
Annotator:
Data collector (2):
Researcher (5):
Supervisor:
-
1.
University of Abou Bekr Belkaïd
- 2. Université M'Hamed Bougara Boumerdès Faculté des Sciences de l'Ingénieur
- 3. LAZOUNI Clinic, Tlemcen, Algeria
- 4. University of Baghdad
-
5.
University of Miami
Description
This dataset contains 1,454 eyes from 744 patients (889 normal, 565 keratoconus) collected with the Orbscan 3 device in Algerian ophthalmology centers.
For each eye, four corneal maps are provided (Axial, Anterior, Posterior, Pachymetry), alongside structured clinical annotations (age, sex, astigmatism, K max, pachymetry, asphericity, diagnostic label).
All data are anonymized and pre-processed into standardized PNG images and CSV files, directly usable for deep learning research.
This resource provides one of the first multimodal Orbscan-based keratoconus datasets from North Africa, supporting AI development in ophthalmology and reproducibility in medical imaging.
Files
ORBSCAN_Dataset.zip
Additional details
Additional titles
- Alternative title (English)
- A Public Multimodal Dataset of Orbscan Topography Images and Clinical Features
Dates
- Updated
-
2026-06-04