The default config file used in topup (${FSLDIR}/src/topup/flirtsch/b02b0.cnf) contains the parameter : --subsamp=2,2,2,2,2,1,1,1,1
This causes an error if the spin echos (and epi) have an odd number of slices (e.g. interlayer data collected by Elyssa and Mark). Quick fix is to supply this parameter to the topup sbatch call with either no subsampling, --subsamp=1,1,1,1,1,1,1,1,1, or a subsampling rate that can divide the number of slices evenly. For example, Elyssa and Mark had 51 slices, so I amended the topup call in sbatch_prep_TOPUP line 185 to:
topup --imain="${all_SE}" --datain="${topup_dir}/acqparams.txt" \
--config=b02b0.cnf --out="${topup_dir}/topup_output" \
--subsamp=3,3,3,3,3,1,1,1,1 \
--iout="${topup_dir}/topup_iout" --fout="${topup_dir}/topup_fout" \
--logout="${topup_dir}/topup_logout"
Ideally, the topup code would automatically do this. I.e. check if the number of slices is even and adjust subsampling accordingly.
The default config file used in topup (
${FSLDIR}/src/topup/flirtsch/b02b0.cnf) contains the parameter :--subsamp=2,2,2,2,2,1,1,1,1This causes an error if the spin echos (and epi) have an odd number of slices (e.g. interlayer data collected by Elyssa and Mark). Quick fix is to supply this parameter to the topup sbatch call with either no subsampling,
--subsamp=1,1,1,1,1,1,1,1,1, or a subsampling rate that can divide the number of slices evenly. For example, Elyssa and Mark had 51 slices, so I amended the topup call insbatch_prep_TOPUPline 185 to:Ideally, the topup code would automatically do this. I.e. check if the number of slices is even and adjust subsampling accordingly.