Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- medical-segmentation
|
| 4 |
+
- pytorch
|
| 5 |
+
- synapse
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
# Synapse Multi-Organ Segmentation (Lightweight U-Net)
|
| 9 |
+
|
| 10 |
+
This model is a lightweight U-Net trained on the Synapse multi-organ CT dataset.
|
| 11 |
+
It is designed to run on low-resource hardware (GTX 1650, 4GB VRAM) and CPU.
|
| 12 |
+
|
| 13 |
+
## Model Details
|
| 14 |
+
- Architecture: U-Net (Filters: 32, 64, 128, 256)
|
| 15 |
+
- Input: 256x256 Axial CT Clices (Windowed [-125, 275], Normalized [0,1])
|
| 16 |
+
- Classes: 9 (Background + 8 organs)
|
| 17 |
+
|
| 18 |
+
## Usage
|
| 19 |
+
Run with the provided `app.py` in the linked GitHub repository.
|
| 20 |
+
|
| 21 |
+
```python
|
| 22 |
+
model = UNet(n_channels=1, n_classes=9)
|
| 23 |
+
model.load_state_dict(torch.load("best_model.pth"))
|
| 24 |
+
```
|
| 25 |
+
|
| 26 |
+
## Dataset
|
| 27 |
+
Synapse Multi-Organ Segmentation Challenge.
|
| 28 |
+
License: Review original dataset license.
|