I am trying to use fsfast to perform 1st level GLM analysis on preprocessed data from the HCP-Aging dataset which happens to be in a different space than fsaverage. I followed the following steps:
First I smoothed the data which was in dtseries format
wb_command -cifti-smoothing *dtseries.nii 1.27 1.27 COLUMN f.sm3.dtseries.nii -left-surface *L.midthickness_MSMAll.32k_fs_LR.surf.gii -right-surface *R.midthickness_MSMAll.32k_fs_LR.surf.gii
Then extracted the left/right surface and volume
wb_command -cifti-separate f.sm3.dtseries.nii COLUMN -metric CORTEX_LEFT f.L.func.gii -metric CORTEX_RIGHT f.R.func.gii -volume-all f.mni152.nii
Followed by creating a mask for the subcortical analysis
mri_binarize --i Atlas_ROIs.2.nii.gz --match 8 10 11 12 13 16 17 18 26 28 47 49 50 51 52 53 54 58 60 --o subcortmask.nii.gz;
and creating a global mean .dat file.
wb_command -cifti-stats f.sm3.dtseries.nii -reduce MEAN > global.waveform.dat
Then I tried to run GLM analysis using FSFast. The subcortical analysis completes without errors, but I am having trouble with the surfaces.
Surprisingly, the error seems to be at the Smoothing ACF step because it is trying to perform it on volume even though I have specified that the analysis should be in surface space. Attached are the selx3 log file and analysis info file. Is there something with the analysis specification or in earlier conversion steps? I would appreciate any help with this.
Thank you
Best,
Pia