Hi FSL gurus:
Can one of you tell me why various fsl routines (eg fslview) would report that a file was filled with "-nan" when in fact there is valid data in it?
I have some nifti files (4D resting data per individual subject) that were processed using CompCor (preprocessing and then nuisance regression, including GSR). The processing code is implemented in SPM and puts out 3D volumes, which I have merged with fslmerge.
When I inspect these in fslview, the initial display shows a min/max of -nan/-nan. However if I reset the min/max values (eg from 3000 to 10000), a brain appears. (However the space outside the brain is -nan, not zero.)
Similarly when I run fslmeants on one of these files, the time courses are output as all zeroes.
The initial raw data looks normal and comes out normally if I preprocess it with the HCP pipeline up through ICA-FIX. If I run FSLinfo on the HCP vs CompCor outputs, it shows the inputs are in the same MNI format.
Would you have an insight as to why the CompCor files are being interpreted wrong?
In case it is helpful, I ran "fslhd" on both the CompCor and ICA-FIX output; below are the discrepancies found when I "diff" the "fslhd" outputs. Two things are a bit puzzling. (1) the pixdim4 should be .72 for both and was correctly set in each pipeline. (2) The two outputs are apparently being interpreted as MNI_152 vs Aligned Anat. Would this cause the problem I have described? Is there an easy way to force the files to be interpreted the same?
diff rfMRI_REST_AP_hp2000_clean_FSLHD_output.txt Merged_CCout_23479_rest_ap_G_FSLHD_output.txt
1c1
< filename rfMRI_REST_AP_hp2000_clean.nii.gz (HCP/ICA-FIX)
---
> filename Merged_CCout_23479_rest_ap_G.nii.gz (CompCor)
22c22
< pixdim4 0.720000
---
> pixdim4 1.000000
46,47c46,47
< qform_name MNI_152
< qform_code 4
---
> qform_name Aligned Anat
> qform_code 2
55,56c55,56
< sform_name MNI_152
< sform_code 4
---
> sform_name Aligned Anat
> sform_code 2
Thanks very much for any insights...
-Christina Fales
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|