Eye Disease Dataset Class Distribution in Excel

Class distribution data for a dataset used to train a MaxViT model for classifying diabetic retinopathy, glaucoma, and cataracts from fundus images. The model achieved 96.75% macro-averaged accuracy. Authored by Md Mehedi Hasan Santo and uploaded to Figshare in April 2026.

Use Cases

• Analyze class balance for diabetic retinopathy, glaucoma, and cataract categories to inform sampling strategies.
• Validate model performance metrics like 96.75% accuracy against the reported class distribution.
• Use the distribution to calculate evaluation metrics weighted by class prevalence for fair model comparison.

Strengths

• Data supports a model achieving 96.75% macro-averaged accuracy on a public benchmark.
• Derived from a study employing rigorous preprocessing and five-fold cross-validation.

Limitations

• The dataset is extremely small at 5.5 KB, containing only summary statistics, not the underlying images or patient records.
• No raw image data, feature columns, or individual sample labels are included, limiting analytical depth.

Provenance

Source

Derived from a publicly available eye disease classification dataset, as cited in the associated research.

Collection Method

Calculated from the dataset used to train and evaluate a multi-axis vision transformer (MaxViT) model.

Freshness

Last updated in April 2026.