https://github.com/JeffFessler/mirt-demo
Collection of Jupyter notebooks that demonstrate the image reconstruction capabilities of MIRT.jl, the Julia language version of the Michigan Image Reconstruction Toolbox.
These demos remain here for historical completeness, but are no longer maintained. Please see the updated examples at https://github.com/JuliaImageRecon/Examples.
You can test drive any of these Jupyter notebooks in a browser without installing any local software by using the free service at https://mybinder.org/ by clicking on the binder badges below:
-
Basic introduction:
-
MRI compressed sensing reconstruction demo:
-
Dynamic MRI with golden-angle radial sampling reconstruction demo:
You can also view the notebook code directly:
Here is an example of the kind of image produced by one of the demos
Some of the notebooks reproduce figures in this survey paper:
- J A Fessler: "Optimization methods for MR image reconstruction," IEEE Sig. Proc. Mag., 37(1):33-40, Jan. 2020.
The Makefile uses some very useful command-line tools,
namely
jupytext
and
nbconvert
to automatically
create the Jupyter notebooks
and a corresponding html version
from the Julia source code.
This software was developed at the University of Michigan by Jeff Fessler and his group.