A collection of deep learning architectures and applications ported to the PyTorch framework and tools for basic medical image processing. ANTsTorch with some cross-compatibility with our Python and R analogs, ANTsPyNet and ANTsRNet, respectively.
ANTsTorch provides several high-level features:
- A large collection of common deep learning architectures for medical imaging that can be initialized
- Various pre-trained deep learning models to perform key medical imaging tasks
- Utility functions to improve training and evaluating of deep learning models on medical images
- Normalizing flows
Installation
Download and install from source:
git clone https://github.com/ANTsX/ANTsTorch
cd ANTsTorch
python -m pip install .
or
python3 -m pip install git+https://github.com/ANTsX/ANTsTorch.git
Applications
ANTsTorch supports several applications (ANTsPyNet weights):
- Data augmentation and preprocessing utilities
- Multi-modal brain extraction
- T1 (brain-only, three-tissue, hemisphere, lobes)
- T2
- T2star
- FA
- FLAIR
- MRA
- Cortical thickness estimation
- Deep Atropos (six-tissue brain segmentation)
- Desikan-Killiany-Tourville cortical labeling
- Harvard-Oxford-Atlas labeling
- Deep FLASH
- Cerebellar morphology
- MRI modality classification
Other ANTsPyNet ports are a WIP.
Publications
See the ANTsX Ecosystem publications for background and applications.
License
The ANTsTorch package is released under an Apache License.
Acknowledgements
- We gratefully acknowledge the grant support of the Office of Naval Research (N0014-23-1-2317).