Synthetic Retinal Fundus Images for Alzheimer's Disease Severity, 30,000 Images

30,000 synthetic retinal fundus images across four Alzheimer's Disease severity stages. Images are 96 by 96 pixels in JPEG format. The dataset is hosted on Kaggle, but author, organization, and creation details are unspecified.

Use Cases

• Train image classification models to detect Alzheimer's Disease severity based on synthetic retinal fundus images.
• Benchmark generative models for medical image synthesis using the provided synthetic fundus images.
• Develop data augmentation pipelines for medical imaging tasks using the 30,000 synthetic samples.
• Study the relationship between retinal features and neurodegenerative disease progression using labeled synthetic data.

Strengths

• Contains 30,000 synthetic images, providing a substantial volume for model training.
• Images are labeled across four distinct Alzheimer's Disease severity stages.
• Images have a consistent resolution of 96 by 96 pixels.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Kaggle

Collection Method

Synthetically generated; specific generation method is unknown.