Hello Teemu,
Can you let me know the results of running fslhd on your input 4D data?
Many Regards
Matthew
> Hi,
>
>
> Feat (4.1.9, OS X 10.6.8) runs ok when I make sure (using the new -tr parameter of fslmerge) that pixdim4 is correct (i.e., =TR).
>
>
> If pixdim4 is not correct (pixdim4!=TR), feat gets into this if clause:
>
> featlib.tcl:
>
> set IMTR [ exec sh -c "$FSLDIR/bin/fslval filtered_func_data pixdim4" ]
>
> if { [ expr abs($IMTR - $fmri(tr)) ] > 0.01 } {
> fsl:exec "${FSLDIR}/bin/fslhd -x filtered_func_data | sed 's/ dt = .*/ dt = '$fmri(tr)'/g' > tmpHeader"
> fsl:exec "${FSLDIR}/bin/fslcreatehd tmpHeader filtered_func_data"
> fsl:exec "/bin/rm tmpHeader"
> }
>
> After this, all resulting files (filtered_func, zstat, thresh_zstat) are full of zeros.
>
> However, feat works ok with incorrect pixdim4 when I first decrease nvols from 1418 to 1300 (the epi file is > 1GB resulting in filtered_func > 2GB).
>
> Is fslcreatehd having difficulties with big files?
>
>
>
> Cheers,
>
> Teemu
>
>
>
> --
> Teemu Rinne
> Institute of Behavioural Sciences
> Univ Helsinki, Finland
> http://www.helsinki.fi/science/atbrain
>
|