|
|
--- |
|
|
license: mit |
|
|
tags: |
|
|
- medical |
|
|
- radiology |
|
|
- image-classification |
|
|
- pytorch |
|
|
- radimagenet |
|
|
- feature-extraction |
|
|
library_name: pytorch |
|
|
--- |
|
|
|
|
|
# RadImageNet Pre-trained Models |
|
|
|
|
|
This repository contains pre-trained models from RadImageNet, a large-scale radiologic image dataset designed to facilitate transfer learning for medical imaging applications. |
|
|
|
|
|
## Model Description |
|
|
|
|
|
RadImageNet models are convolutional neural networks pre-trained on a diverse collection of radiologic images spanning multiple modalities and anatomical regions. These models serve as powerful feature extractors for downstream medical imaging tasks. |
|
|
|
|
|
### Available Models |
|
|
|
|
|
- **ResNet50.pt**: ResNet-50 architecture pre-trained on RadImageNet |
|
|
- **DenseNet121.pt**: DenseNet-121 architecture pre-trained on RadImageNet |
|
|
- **InceptionV3.pt**: Inception-V3 architecture pre-trained on RadImageNet |
|
|
|
|
|
## Usage |
|
|
|
|
|
```python |
|
|
import torch |
|
|
from huggingface_hub import hf_hub_download |
|
|
|
|
|
# Download and load a model |
|
|
model_path = hf_hub_download(repo_id="Lab-Rasool/RadImageNet", filename="ResNet50.pt") |
|
|
model = torch.load(model_path, map_location="cuda" if torch.cuda.is_available() else "cpu") |
|
|
model.eval() |
|
|
|
|
|
# Use for inference |
|
|
# ... your inference code here ... |
|
|
``` |
|
|
|
|
|
## Preprocessing |
|
|
|
|
|
Images should be preprocessed using standard ImageNet normalization: |
|
|
|
|
|
```python |
|
|
from torchvision import transforms |
|
|
|
|
|
preprocess = transforms.Compose([ |
|
|
transforms.Resize(256), |
|
|
transforms.CenterCrop(224), |
|
|
transforms.ToTensor(), |
|
|
transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]) |
|
|
]) |
|
|
``` |
|
|
|
|
|
## Citation |
|
|
|
|
|
If you use these models in your research, please cite the RadImageNet paper: |
|
|
|
|
|
```bibtex |
|
|
@article{mei2022radimagenet, |
|
|
title={RadImageNet: An Open Radiologic Deep Learning Research Dataset for Effective Transfer Learning}, |
|
|
author={Mei, Xueyan and Liu, Zelong and Robson, Philip M and Marinelli, Brett and Huang, Mingqian and Doshi, Amish and Jacobi, Adam and Cao, Chendi and Link, Katherine E and Yang, Thomas and others}, |
|
|
journal={Radiology: Artificial Intelligence}, |
|
|
volume={4}, |
|
|
number={5}, |
|
|
pages={e210315}, |
|
|
year={2022}, |
|
|
publisher={Radiological Society of North America} |
|
|
} |
|
|
``` |
|
|
|
|
|
## License |
|
|
|
|
|
MIT License |
|
|
|
|
|
Copyright (c) 2021 BMEII-AI |
|
|
|
|
|
Permission is hereby granted, free of charge, to any person obtaining a copy |
|
|
of this software and associated documentation files (the "Software"), to deal |
|
|
in the Software without restriction, including without limitation the rights |
|
|
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell |
|
|
copies of the Software, and to permit persons to whom the Software is |
|
|
furnished to do so, subject to the following conditions: |
|
|
|
|
|
The above copyright notice and this permission notice shall be included in all |
|
|
copies or substantial portions of the Software. |
|
|
|
|
|
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
|
|
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
|
|
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
|
|
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
|
|
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
|
|
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE |
|
|
SOFTWARE. |
|
|
|
|
|
## Additional Information |
|
|
|
|
|
- **Original Repository**: [BMEII-AI/RadImageNet](https://github.com/BMEII-AI/RadImageNet) |
|
|
- **Paper**: [RadImageNet: An Open Radiologic Deep Learning Research Dataset](https://pubs.rsna.org/doi/10.1148/ryai.210315) |
|
|
- **Dataset**: The RadImageNet dataset contains 1.35 million annotated radiologic images |
|
|
|
