Reference implementation of:
Joint Design of RF and Gradient Waveforms via Auto-Differentiation for 3D Tailored Exitation in MRI

cite as:

@misc{luo2020joint,
  title={Joint Design of RF and gradient waveforms via auto-differentiation for 3D tailored excitation in MRI},
  author={Tianrui Luo and Douglas C. Noll and Jeffrey A. Fessler and Jon-Fredrik Nielsen},
  year={2020},
  eprint={2008.10594},
  archivePrefix={arXiv},
  primaryClass={eess.IV},
  url={https://arxiv.org/abs/2008.10594}
}

For the interpT feature, consider citing:

@inproceedings{luo2021MultiScale,
  title={Multi-scale Accelerated Auto-differentiable Bloch-simulation based joint design of excitation RF and gradient waveforms},
  booktitle={ISMRM},
  pages={0000},
  author={Tianrui Luo and Douglas C. Noll and Jeffrey A. Fessler and Jon-Fredrik Nielsen},
  year={2021}
}

• Ubuntu 18.04, Python 3

The implementation may fail with other configurations.

setup_AutoDiffPulses.m does the configurations for Matlab.
For the python part, in your command line, navigate to the repo's root directory, type:

pip install .

Demos are provided in ./demo.

This repo has included binary test data files for basic accessibility in certain regions.
Future binary data files will be added to: https://drive.google.com/drive/folders/1EyKhA_d74OC4KADMuTd1kRTEMoVqWdIY.

This work requries Python (≥v3.5), PyTorch (≥v1.3) with CUDA.

• MRphy: Python, Github link (≥v0.1.5).
• +mrphy: Matlab, Github link.
• +attr: Matlab, Github link.

Other Python dependencies include:
scipy, numpy, PyTorch.