Chákṣu IMAGE: Glaucoma Fundus Images for Indian Ethnicity

Chákṣu IMAGE is a database of 1,345 retinal fundus images for glaucoma prescreening, created through a collaboration between the Indian Institute of Science and Manipal Academy of Higher Education. The database includes images from three different camera devices and provides ground-truth annotations for the optic disc, optic cup, and a binary glaucoma diagnosis from five expert ophthalmologists. It is described as the largest Indian ethnicity-specific fundus image database.

Use Cases

• Training segmentation models for the optic disc and optic cup based on expert-provided ground-truth.
• Developing binary classification models for glaucoma detection based on the normal/glaucomatous labels.
• Benchmarking model performance across different fundus camera devices mentioned in the description.
• Studying glaucoma presentation and model generalizability within a specific Indian demographic.

Strengths

• Contains 1,345 annotated retinal fundus images.
• Provides ground-truth from five expert ophthalmologists for segmentation and diagnosis.
• Includes images from three distinct camera devices, covering resolutions from 1920x1440 to 2448x3264 pixels.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.
• Data may reflect geographic and demographic bias inherent to its specific focus on Indian ethnicity.

Provenance

Source

Indian Institute of Science (IISc) and Manipal Academy of Higher Education (MAHE)

Collection Method

Retinal color fundus images were acquired using three different camera devices from clinical settings.

Geography

India (specific to Indian ethnicity)