ISMRM 2019 Reproducible research study group challenge

HTML pages : https://atrotier.github.io/ISMRM_RRSG_JULIA/Julia_CG-sense.html generated with the static-export-template workflow for Pluto

Description

Implementation of ISMRM Reproducibility Challenges using the MRI reconstruction framework MRIReco.jl.

This implentation uses low-level function of multiple packages and implement the conjugate gradient algorithm in a plain text function. The implementation is done in a Pluto Notebook.

Challenge 1

Goal of challenge 1 is to reproduce the results from the paper [1].

Results

The following figures show the reconstruction results generated with Julia_CG-sense.jl. In the left figure, Δ denotes the NRMSE between the iterates and the reference reconstruction. In the right figure, δ denotes the ratio of the cost function at each iteration to the initial cost function.

Figure 4 : Error results after 10 iteration


Figure 5 : Brain reconstruction after 10 iterations


References

[1] Pruessmann, K. P.; Weiger, M.; Boernert, P. and Boesiger, P. Advances in sensitivity encoding with arbitrary k-space trajectories. Magn Reson Med 46: 638-651 (2001)

Contact

Aurélien Trotier [aurelien . trotier (at) rmsb.u-bordeaux . fr]