Hi all,
I used dwi2response dhollander to estimate the response function. But, in the running, the program print the following information to me. It is said ‘tournier’ algorithm is running. Does this make sense?
dwi2response: eRunning ‘tournier’ algorithm to select 1563 single-fibre WM voxels.e
edwi2response tournier dwi.mif _respsfwmss.txt -sf_voxels 1563 -iter_voxels 15630 -mask refined_wm.mif -voxels voxels_sfwm.mif -tempdir /dwiDir
Best,
Zonglei
Hi @Zonglei,
No worries, that’s entirely intended behaviour: the dhollander algorithm uses the tournier one to determine the single-fibre white matter voxels. The difference in running tournier outright with defaults for that task, is that the dhollander algorithm initialises the tournier algorithm specifically with a rough mask that already filtered out a lot of GM and CSF at that stage. On top of that, it also asks the tournier algorithm for a specific number of voxels, as a very small percentage of what was still left in the WM mask at that stage. In that way, the whole construct deals nicely with widely varying spatial resolutions, as well as unusual situations with a lot of atrophy. In also avoids lesions, which by this stage of the dhollander algorithm are mostly not a part any more of any of the masks. See the figure on the right in the abstract (as well as the description of the 4th step somewhere in the text) for more insight!
Cheers,
Thijs
I see, thanks for your kind interpretation. @ThijsDhollander