|
|
--- |
|
|
title: Spleen Segmentation Demo |
|
|
emoji: 🖥️ |
|
|
colorFrom: blue |
|
|
colorTo: gray |
|
|
sdk: gradio |
|
|
sdk_version: 5.49.0 |
|
|
app_file: app.py |
|
|
pinned: false |
|
|
short_description: 3D spleen segmentation with MONAI |
|
|
models: |
|
|
- MONAI/example_spleen_segmentation |
|
|
--- |
|
|
|
|
|
# CT Spleen Segmentation Demo |
|
|
|
|
|
This Space demonstrates 3D spleen segmentation from CT scans using the [MONAI/example_spleen_segmentation](https://huggingface.co/MONAI/example_spleen_segmentation) model. |
|
|
|
|
|
## Model Information |
|
|
|
|
|
- **Architecture**: UNet |
|
|
- **Input**: 3D CT images (96×96×96) |
|
|
- **Output**: Binary segmentation (spleen vs background) |
|
|
- **Performance**: Mean Dice Score = 0.96 |
|
|
- **Training**: Trained on Medical Segmentation Decathlon Challenge 2018 dataset |
|
|
|
|
|
## How to Use |
|
|
|
|
|
1. Upload a CT scan in NIfTI format (.nii or .nii.gz) |
|
|
2. Click "Segment Spleen" |
|
|
3. View the segmentation overlay (middle slice visualization) |
|
|
4. Download the full 3D segmentation |
|
|
|
|
|
## Requirements |
|
|
|
|
|
- MONAI |
|
|
- PyTorch |
|
|
- nibabel |
|
|
- numpy |
|
|
- huggingface_hub |
|
|
|
|
|
## Citation |
|
|
|
|
|
If you use this model, please cite: |
|
|
|
|
|
``` |
|
|
Xia, Yingda, et al. "3D Semi-Supervised Learning with Uncertainty-Aware Multi-View Co-Training." |
|
|
arXiv preprint arXiv:1811.12506 (2018). |
|
|
``` |
|
|
|
|
|
## Disclaimer |
|
|
|
|
|
This is an example demonstration, not to be used for diagnostic purposes. |
