Autofluorescence Intensity Patterns for Alpha and Beta Cell Identity in Human Islets

A 16.5 GB repository of pre-processed image frames and trained models for a study on autofluorescence patterns in human pancreatic islets. The data and code were authored by Ivano Squiccimarro and last updated in May 2026. The repository includes raw data, pre-trained machine learning models, and comprehensive analysis results.

Use Cases

• Training image classification models based on autofluorescence intensity patterns.
• Reproducing the study's analysis of alpha and beta cell identity encoding.
• Comparing machine learning model performance using the provided supportive datasets.
• Applying interpretability methods to the pre-trained models for biological insight.

Strengths

• Includes pre-trained machine learning models, reducing initial training overhead.
• Large 16.5 GB scale suggests substantial underlying image data.
• Associated source code is publicly available on GitHub for full reproducibility.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect bias inherent to the specific experimental and imaging methodology.

Provenance

Source

figshare

Collection Method

Likely generated from experimental imaging of human pancreatic islets.

Time Range

null

Freshness

Last updated 2026-05-04 15:24:16; freshness should be verified.

Geography

null