Dear all,
I’m attempting to run the rsfMRI dHCP pipeline on our own data. Unfortunately, we don’t have a fieldmap, and I am currently unable to run the pipeline. Is there a workaround?
Many thanks in advance!
Best wishes,
Fem
Dear all,
I’m attempting to run the rsfMRI dHCP pipeline on our own data. Unfortunately, we don’t have a fieldmap, and I am currently unable to run the pipeline. Is there a workaround?
Many thanks in advance!
Best wishes,
Fem
Hi Fem
The dHCP fMRI pipeline is very dependent on having a field map. It would, in principle, be possible to run without a field map but would require tweaking of the code.
If you have collected diffusion data, with two B0’s with opposing phase-encode direction, then the pipeline will work with that?
If you have not, then I will have to evaluate how extensive the changes required to support running without field maps are.
Cheers, Sean
Many thanks for the prompt reply, Sean!
Yes, I was thinking about using the DTI as well. But it’s older data, and we don’t have B0’s with AP/PA phase-encoding.
It would be great if you could evaluate the feasibility of changing the code. Or maybe you have recommendations of using a different pipeline/tool?
Best wishes,
Fem
Dear Sean,
Just touching base for a quick update on this issue. If you could let me know whether it’s possible to change the code, would be appreciated!
Best wishes, Femke
Hi Femke
Sorry for the delay. I have started making the necessary mods. I don’t think it will be too onerous and should be able to give you something to test later this week.
Cheers, Sean
Hi Femke
Sorry for the delay. It turns out the changes to run without a field map were straight forward at the lower levels of the pipeline as I expected, but it was more challenging than I expected to incorporate this into the higher-level code and maintain backward compatibility.
Therefore, I have instead written two example scripts to demonstrate how the pipeline can be run at a lower-level.
The first script demonstrates the default dHCP pipeline steps: https://git.fmrib.ox.ac.uk/seanf/dhcp-neonatal-fmri-pipeline/-/blob/without_fieldmap/dhcp/examples/dhcp-default.py
The second script demonstrates running without a fieldmap: https://git.fmrib.ox.ac.uk/seanf/dhcp-neonatal-fmri-pipeline/-/blob/without_fieldmap/dhcp/examples/dhcp-no-fieldmap.py
I have not committed this to the master branch yet. They are available on a branch called “without_fieldmap”. I will merge to the master branch when I have finished testing.
Again, sorry for the delay. I hope this helps.
Cheers, Sean
Dear Sean,
Thank you very much for implementing these changes! Amazing!
Best wishes,
Femke
No problem. The without_fieldmap branch has been merged into the master branch.
Cheers, Sean
Dear Sean,
Quick question. In the eddy-correction phase it generates the error: EddyInputError: --b0_only disabled awaiting release of eddy for fmri (freddy). Maybe I’m using the wrong fsl/eddy version? A google search doesn’t show any results for “freddy” or “b0_only”.
Any suggestions?
Best wishes, Femke
Hi Femke
The pipeline requires FSL 6.04. It sounds like you are using an earlier version.
You can check your installed version with this command:
cat $FSLDIR/etc/fslversion
Cheers, Sean
MANAGED BY INCF