Official code for "MINeR: Direction-modulated implicit neural representation enables ultrafast multi-shell diffusion MRI"
This repo holds the codes for MINeR.
To run the reconstruction demo, the following dependencies are required:
- Python 3.10
- PyTorch 2.0.0
- tinycudann 1.7
- numpy 1.22.4
- nibabel 5.3.2
- (optional) matplotlib, tensorboard, h5py
This repo provides a toy undersampled demo dataset in demo_data/.
Full datasets can be downloaded from the HCP (Human Connectome Project) database.
python main.py --mode train --demo_dir demo_data --output_dir outputs
python main.py --mode infer --demo_dir demo_data --checkpoint outputs/model.pt --output_dir outputs