Single-Cell Pathology Data for Breast Cancer with Images and Cell Masks

High-dimension TIFF images, single-cell and tumor/stroma masks, and single-cell patient data support the findings of the publication "The Single-Cell Pathology Landscape of Breast Cancer". The dataset, associated with the University of Zurich, includes spillover-corrected single-cell data for downstream R analysis. Original mask files contain labels from CellProfiler, while exported data uses sequential labels from histoCAT.

Use Cases

• Train image segmentation models for tumor and stromal regions based on provided masks.
• Analyze spatial relationships between cell types in breast cancer tissue based on single-cell masks and coordinates.
• Conduct downstream single-cell analysis on spillover-corrected protein expression or morphological features.
• Validate computational pathology pipelines using the provided high-resolution TIFF images and ground truth annotations.

Strengths

• Includes multiple aligned data modalities: high-resolution images, single-cell masks, and tumor/stroma region masks.
• Single-cell data is pre-processed and spillover-corrected, ready for downstream analysis.
• Provides both original CellProfiler mask labels and sequential histoCAT labels for data matching.

Limitations

• Row count and dataset size are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

University of Zurich, associated with the BodenmillerGroup GitHub repository.

Collection Method

Data generated and published as part of a research study on single-cell pathology.