A repository that contains reconstruction methods for 2D spiral 129-Xe data, with inclusion of flip angle and N4 corrections.
- MATLAB simulation scripts can be found in the
matlab\folder. - Python in vivo scripts can be found in the
python\folder. It is recommended that you set up the Python virtual environment using the instructions below.
Please post issues to the GitHub repository if you have any questions or improvements.
Written by Joseph Plummer. joseph.plummer@cchmc.org
Co-authors:
- Riaz Hussain
- Abdullah Bdaiwi
- Matthew Willmering
- Mariah Costa
- Juan Parra-Robles
- Zackary Cleveland
- Laura Walkup
Run the following commands in sequence to run the experiments.
conda update -n base -c defaults condamake condaconda activate spiral-reconmake pip
Troubleshooting:
- This repository was tested on an NVIDIA GPU. If running on a system without
the same, please remove the following packages from
environment.yaml:cudnnncclcupy
- Additionally, if not using an NVIDIA GPU, please set
devnum = -1for each of the*.pyfiles. - When running
make pip,pip install git+https://github.com/mikgroup/sigpy.git@mainwill error if GitHub ssh keys are not set. If this is the case, install SigPy manually.
Sample multi-slice 2D spiral data collected from a healthy participant included inside folder called data.
Run the scripts inside matlab and python folders. I recommend using VSCODE with WSL2 for the Python script.
To uninstall, run the following commands:
conda activatemake clean