Diagnostic Performance of AI Systems for Glaucoma Detection from Fundus Photos

A retrospective cross-sectional study from 2026 evaluates the diagnostic accuracy of two AI systems for glaucoma detection using color fundus photography. The dataset includes results from 370 eyes of 193 patients, comparing the Laguna ONhE and VUNO Med-Fundus AI systems against structural and functional tests. It was authored by Gilvan Vilarinho da Silva Filho and shared under a CC-BY-4.0 license.

Use Cases

• Benchmarking AI diagnostic models for glaucoma based on reported AUC metrics.
• Studying the combined performance of multiple AI systems as mentioned in the description.
• Developing screening tools for resource-limited settings based on the use of color fundus photography alone.

Strengths

• Includes results from a specific cohort of 370 eyes from 193 patients.
• Reports concrete diagnostic performance metrics, including AUC values of 0.879 and 0.857 for individual AI systems.
• Data is shared under a permissive CC-BY-4.0 license.

Limitations

• Row count and column-level documentation are unknown, limiting suitability assessment.
• The dataset is very small at 5.5 KB, indicating limited scope.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare

Collection Method

Retrospective cross-sectional study analyzing fundus photographs with AI systems.

Freshness

Last updated 2026-05-28 17:34:37; freshness should be verified.