# Paired brightfield and fluorescence images of mouse brain and dura mater for virtual immunostaining (DAPI, IBA1, CD206, LYVE1, neutrophil)

## Facts

- Provider: Yibin Liu et al.
- License: CC-BY-4.0 (https://huggingface.co/datasets/YibinLiu/mouse-brain-dura-virtual-staining)
- Size: 4 files
- Stored at: Hugging Face, https://huggingface.co/datasets/YibinLiu/mouse-brain-dura-virtual-staining
- Access: public, readable without an account
- Category: Neuroscience
- Source: https://huggingface.co/datasets/YibinLiu/mouse-brain-dura-virtual-staining
- Page: https://scigantic.com/archive/56fa2705-d3a1-5c14-b200-1a5bcd55ebd6
- Keywords: Neuroscience, benchmark, Hugging Face

## About

Paired brightfield and immunofluorescence microscopy images of brain and dura mater from 19 male C57BL/6J mice aged 6 to 8 weeks, treated with phosphate-buffered saline (PBS) or lipopolysaccharide (LPS) to model neuroinflammation. The release holds 4,526 original TIFF files (1,335 brightfield fields and 3,191 task-specific fluorescence pairs) for five targets: DAPI nuclei, IBA1 microglia and macrophages, CD206 (mannose receptor), LYVE1 lymphatic and perivascular macrophages, and a neutrophil-associated antibody channel. Brightfield was acquired from the same field after immunostaining. For deep learning there are 70,906 paired 256x256 patches stored as float32 NPY arrays with per-task metadata.csv files and a frozen animal-level train/val/test split (51,138 / 7,704 / 12,064 patches), plus preprocessing reports, the training and evaluation code, and 75 benchmark run reports (five models, five targets, three seeds). Brain specimens come from 6 animals and dura from 13. Everything ships as one 39 GB zip (mouse_brain_dura_virtual_staining_v1.0.0.zip) with SHA256SUMS.txt; read TIFFs with tifffile and patches with numpy. Use for virtual staining (brightfield to fluorescence image translation) and meningeal immune cell studies. Version 1.0.0; the authors state a data license of CC BY 4.0 and ask users to cite the forthcoming article and dataset record.

## Load the data

```python
from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="YibinLiu/mouse-brain-dura-virtual-staining", filename="mouse_brain_dura_virtual_staining_v1.0.0.zip", repo_type="dataset")
print(path)
```

## Data structure

Example file: `mouse_brain_dura_virtual_staining_v1.0.0.zip`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/56fa2705-d3a1-5c14-b200-1a5bcd55ebd6
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/56fa2705-d3a1-5c14-b200-1a5bcd55ebd6
- Over MCP: `get_schema_card` with id `56fa2705-d3a1-5c14-b200-1a5bcd55ebd6` at https://api.scigantic.com/mcp

## Related topics

- [Neuroscience datasets](https://scigantic.com/topics/neuroscience)
- [Benchmark datasets](https://scigantic.com/topics/benchmark)
