189,744 labeled nuclei with instance segmentation masks across 19 tissue types and 5 cell categories. The dataset contains 7,901 images captured at x40 magnification, each sized 256x256 pixels with a resolution of 0.25 µm/pixel. It was created by RationAI and last updated on Hugging Face in December 2025.
Use Cases
- Training instance segmentation models based on the 189,744 labeled nuclei masks.
- Developing multi-class nuclei classifiers based on the 5 distinct cell categories.
- Benchmarking model performance across diverse tissue types based on the 19 included types.
- Researching nuclei morphology and distribution in histopathology images based on the x40 magnification and resolution.
Strengths
- Contains 189,744 labeled nuclei, providing a substantial volume of annotations.
- Covers 19 tissue types, suggesting diversity in the medical image domain.
- Includes 7,901 images at a consistent 256x256 pixel size and 0.25 µm/pixel resolution.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- The description notes the dataset is semi-automatically generated, which may imply annotation quality variance.
Provenance
- Source
- RationAI
- Collection Method
- Semi-automatically generated, according to the description.
- Time Range
- null
- Freshness
- Last updated 2025-12 05 13:56:04; freshness should be verified.
- Geography
- null