Brain Region Segmentation — Human Brain (BigBrain Tissue Classification)
DINOv2-Large + UperNet model fine-tuned for semantic segmentation of human brain tissue types in histological sections.
Model Details
| Attribute | Value |
|---|---|
| Architecture | DINOv2-Large (304M) + UperNet (38M) |
| Classes | 10 (tissue types) |
| Input Size | 518x518 |
| Training Data | BigBrain 3D histological volume (200um, 9-class tissue classification) |
| mIoU (center-crop) | 60.8% |
| mIoU (sliding window) | 61.3% |
Tissue Classes
| ID | Class |
|---|---|
| 0 | Background |
| 1 | Gray Matter |
| 2 | White Matter |
| 3 | Cerebrospinal Fluid |
| 4 | Meninges |
| 5 | Blood Vessels |
| 6 | Bone/Skull |
| 7 | Muscle |
| 8 | Artifact |
| 9 | Other/Unknown |
Usage
git clone https://github.com/Noel-Niko/histological-image-analysis
cd histological-image-analysis
make install
make download-models-human-bigbrain
make annotate-human-bigbrain IMAGES=/path/to/your/slides/
Paper
Cross-Species Transfer of Ultra-Fine-Grained Brain Segmentation: From Mouse to Human with DINOv2 + UperNet
This model is Track B of a three-track human brain segmentation study. It uses the BigBrain 200um classified volume with dense 9-class tissue annotations (Merker stain). The BigBrain model serves as a tissue type classifier — complementary to the Allen depth-3 model's role as a brain region identifier.
See paper.md in this repo for the full paper.
Citation
If you use this model, please cite the training data sources and the paper included in this repository.
Repository
Full source code, training notebooks, and all models: https://github.com/Noel-Niko/histological-image-analysis
Maintaining This Repo
To update model weights, papers, or this README:
cd histological-image-analysis
export HUGGING_FACE_TOKEN=hf_your_token_here
# Update model weights (Databricks or local):
jupyter notebook notebooks/upload_models_to_hf.ipynb
# Update papers + READMEs (local only):
jupyter notebook notebooks/upload_papers_to_hf.ipynb
- Downloads last month
- 51