Upload SynthStroke baseline model
Browse files- README.md +85 -0
- model.safetensors +3 -0
README.md
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: pytorch
|
| 4 |
+
tags:
|
| 5 |
+
- medical
|
| 6 |
+
- segmentation
|
| 7 |
+
- stroke
|
| 8 |
+
- neurology
|
| 9 |
+
- mri
|
| 10 |
+
pipeline_tag: image-segmentation
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Baseline
|
| 14 |
+
|
| 15 |
+
Baseline model trained on real ATLAS T1w images.
|
| 16 |
+
|
| 17 |
+
## Model Details
|
| 18 |
+
|
| 19 |
+
- **Name**: Baseline
|
| 20 |
+
- **Classes**: 0 (Background), 1 (Stroke)
|
| 21 |
+
- **Patch Size**: 192³
|
| 22 |
+
- **Voxel Spacing**: 1mm³
|
| 23 |
+
- **Input Channels**: 1 (T1w MRI)
|
| 24 |
+
|
| 25 |
+
## Usage
|
| 26 |
+
|
| 27 |
+
### Loading from Hugging Face Hub
|
| 28 |
+
|
| 29 |
+
```python
|
| 30 |
+
import torch
|
| 31 |
+
from synthstroke_model import SynthStrokeModel
|
| 32 |
+
|
| 33 |
+
# Load the model from Hugging Face Hub
|
| 34 |
+
model = SynthStrokeModel.from_pretrained("liamchalcroft/synthstroke-baseline")
|
| 35 |
+
|
| 36 |
+
# Prepare your input (example shape: batch_size=1, channels=1, H, W, D)
|
| 37 |
+
input_tensor = torch.randn(1, 1, 192, 192, 192)
|
| 38 |
+
|
| 39 |
+
# Get predictions (with optional TTA for improved accuracy)
|
| 40 |
+
predictions = model.predict_segmentation(input_tensor, use_tta=True)
|
| 41 |
+
|
| 42 |
+
# Get lesion probability map (channel 1)
|
| 43 |
+
lesion_probs = predictions[:, 1] # Shape: (batch_size, H, W, D)
|
| 44 |
+
|
| 45 |
+
# Alternative: Get logits without TTA
|
| 46 |
+
logits = model.predict_segmentation(input_tensor, apply_softmax=False)
|
| 47 |
+
```
|
| 48 |
+
|
| 49 |
+
## Citation
|
| 50 |
+
|
| 51 |
+
```bibtex
|
| 52 |
+
@article{chalcroft2025synthetic,
|
| 53 |
+
title={Synthetic Data for Robust Stroke Segmentation},
|
| 54 |
+
author={Chalcroft, Liam and Pappas, Ioannis and Price, Cathy J. and Ashburner, John},
|
| 55 |
+
journal={Machine Learning for Biomedical Imaging},
|
| 56 |
+
volume={3},
|
| 57 |
+
pages={317--346},
|
| 58 |
+
year={2025},
|
| 59 |
+
publisher={Machine Learning for Biomedical Imaging},
|
| 60 |
+
doi={10.59275/j.melba.2025-f3g6},
|
| 61 |
+
url={https://www.melba-journal.org/papers/2025:014.html}
|
| 62 |
+
}
|
| 63 |
+
```
|
| 64 |
+
|
| 65 |
+
For the original arXiv preprint:
|
| 66 |
+
|
| 67 |
+
```bibtex
|
| 68 |
+
@article{Chalcroft_2025,
|
| 69 |
+
title={Synthetic Data for Robust Stroke Segmentation},
|
| 70 |
+
volume={3},
|
| 71 |
+
ISSN={2766-905X},
|
| 72 |
+
url={http://dx.doi.org/10.59275/j.melba.2025-f3g6},
|
| 73 |
+
DOI={10.59275/j.melba.2025-f3g6},
|
| 74 |
+
number={August 2025},
|
| 75 |
+
journal={Machine Learning for Biomedical Imaging},
|
| 76 |
+
publisher={Machine Learning for Biomedical Imaging},
|
| 77 |
+
author={Chalcroft, Liam and Pappas, Ioannis and Price, Cathy J. and Ashburner, John},
|
| 78 |
+
year={2025},
|
| 79 |
+
month=aug, pages={317–346}
|
| 80 |
+
}
|
| 81 |
+
```
|
| 82 |
+
|
| 83 |
+
## License
|
| 84 |
+
|
| 85 |
+
MIT License - see the [LICENSE](https://github.com/liamchalcroft/synthstroke/blob/main/LICENSE) file for details.
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d56c089e8c4bcc0ad2281f1e80b7c0e265f3b7138dee17fb2d160487604eee66
|
| 3 |
+
size 74468100
|