Hi Jesper,
I'm getting a similar issue to Hamad when I run eddy_cuda:
This is the eddy command I'm using:
eddy_cuda9.1 --imain=rawdataAPdn.nii --mask=binary_mask.nii.gz --acqp=acqparams.txt --index=index.txt --bvecs=bvecs --bvals=bvals --topup=topup_results --data_is_shelled --niter=8 --fwhm=10,8,4,2,0,0,0,0 --repol --mporder=32 --s2v_niter=10 --s2v_lambda=5 --s2v_interp=trilinear --slspec=slspec.txt --out=eddycorr
I suspect this is because the acquisition for the data is run with 5 b-value shells with low number of directions for each (b=200 shell only has 9 directions for example). But when I run this using non-cuda eddy (without s2v correction):
eddy_openmp --imain=rawdataAPdn.nii --mask=binary_mask.nii.gz --acqp=acqparams.txt --index=index.txt --bvecs=bvecs --bvals=bvals --topup=topup_results --data_is_shelled --mb=2 --repol --out=eddycorr
There are no errors. I have tried increasing the ol_nstd to 7-8 which seems to work for the data but of course this means I could be missing more outliers than I would otherwise. Why is there a discrepancy here between eddy_openmp and eddy_cuda?
Thanks in advance for your help.
Nikitas
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|