Skip to content

Commit

Permalink
add masking commets
Browse files Browse the repository at this point in the history
  • Loading branch information
korbinian90 committed Oct 10, 2023
1 parent 52ac6b8 commit 371d44a
Show file tree
Hide file tree
Showing 2 changed files with 15 additions and 1 deletion.
14 changes: 14 additions & 0 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -140,3 +140,17 @@ For Mac GPU:
using Metal
chi = qsm_tgv(phase, mask, res; TE, fieldstrength, gpu=Metal);
```

## Masking for QSM

Masking for QSM is a challenge, since including corrupted phase areas can cause global artefacts in the QSM result. See the publications [QSMxT](https://doi.org/10.1002%2Fmrm.29048) and [phase based masking](https://doi.org/10.1002/mrm.29368).

A simple solution is to remove areas based on phase quality using [ROMEO](https://onlinelibrary.wiley.com/doi/10.1002/mrm.28563)

```julia
using MriResearchTools
mask_phase = robustmask(romeovoxelquality(phase; mag))
mask_combined = mask_phase .& mask_brain
```

The mask might contain holes, and a more sophisticated approach is taken in the [QSMxT toolbox](https://qsmxt.github.io/QSMxT/).
2 changes: 1 addition & 1 deletion src/tgv.jl
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,7 @@ function qsm_tgv(phase, mask, res; TE, B0_dir=[0, 0, 1], fieldstrength=3, regula
ret = nothing
end

@showprogress dt=0.5 desc="Running $max_iterations TGV iterations..." for k in 1:max_iterations
@showprogress dt=1 desc="Running $max_iterations TGV iterations..." for k in 1:max_iterations
#############
# dual update
KernelAbstractions.synchronize(device)
Expand Down

0 comments on commit 371d44a

Please sign in to comment.