BBBC038: 2018 Data Science Bowl Nuclei Segmentation Benchmark

A 2D light-microscopy dataset for cell-nucleus segmentation assembled across many imaging experiments. The collection spans multiple modalities including fluorescence and brightfield histopathology, covering humans, mice, and flies across 30+ experiments. This is the official BBBC038v1 release from the Broad Bioimage Benchmark Collection, hosted on Hugging Face by MedOtter.

Use Cases

• Training nuclei segmentation models based on 2D light-microscopy images.
• Benchmarking algorithm performance across different imaging modalities like fluorescence and brightfield.
• Developing cross-modality segmentation techniques based on the dataset's deliberate inclusion of multiple modalities.
• Studying cell nucleus morphology across different species and cell types mentioned in the description.

Strengths

• Deliberately spans multiple imaging modalities including fluorescence (DAPI/Hoechst) and brightfield H&E histopathology.
• Assembled across many experiments, covering 22 cell types and 15 resolutions.
• Serves as a standard cross-modality nuclei-segmentation benchmark in the field.

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

Broad Bioimage Benchmark Collection (BBBC038v1), hosted by MedOtter on Hugging Face.

Collection Method

Assembled from many light-microscopy imaging experiments.

Freshness

Last updated 2026-06-27 00:43:41; freshness should be verified.