Hi MRTrix community,
I am using dwifslpreproc to preprocess some neonatal DWI scans and I don’t have access to GPUs so am running eddy_openmp. On the FSL mailing list it says that the FSL6.0.5 version of eddy_openmp (changed to eddy_cpu in the latest versions) should support slice-to-volume motion correction. I have FSL6.0.7 installed but trying to run eddy from dwifslpreproc with any of the slice-to-volume options eg
dwifslpreproc "${fname_concat}" "${fname_eddy}" -pe_dir AP -rpe_all -force \
-fslgrad "${fname_bvec}" "${fname_bval}" \
-eddy_slspec "${fname_slspec}" \
-eddyqc_all "${dirname_eddyqc}" \
-nocleanup \
-scratch "${dirname_scratch}" \
-continue "${dirname_scratch}" eddy_in.nii \
-eddy_options " --residuals --cnr_maps --repol --slm=linear --mporder=5 --s2v_niter=3"
gives the error:
EddyInputError: Slice-to-vol is not yet implemented for the CPU version.
Terminating program
Has anyone tried to run the slice-to-vol correction with the CPU version? Is it unbearably slow? And is there any plans to incorporate it into dwifslpreproc?
Thanks!
Diliana