Dataset Composition and Splits for a Medical Imaging Study

A summary of dataset composition and data splits used in a study of diagnostic performance metrics. The dataset was created by Nisreen Osman E. Ahmed and last updated on March 17, 2026. It is a 5.5 KB Excel file describing the data used for model training and evaluation.

Use Cases

• Replicating the study's experimental setup based on the described data splits.
• Analyzing the impact of data augmentation strategies based on the description of its exclusive use during training.
• Benchmarking diagnostic performance metrics based on the use of raw, non-augmented slides for evaluation.
• Studying calibration techniques like temperature scaling and isotonic regression based on the platform tags.

Strengths

• Clear license (CC-BY-4.0) permitting reuse with attribution.
• Specific file size of 5.5 KB indicates a concise, focused metadata summary.
• Explicitly states that diagnostic metrics were computed on raw, non-augmented slides, clarifying the evaluation protocol.

Limitations

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

Provenance

Source

figshare, authored by Nisreen Osman E. Ahmed.

Collection Method

Likely compiled as part of a research study on medical image analysis and model calibration.

Time Range

null

Freshness

Last updated 2026-03-17 17:37:42; freshness should be verified.

Geography

null