Dear All,
I have outlined the exact steps within our pipeline below. When fitting any glm to the data (simple osgm shown below), we find that across all vertices, the significance values are 0 (despite the vertex by vertex diff values appearing to be very slightly different, based on a cursory inspection of the sig.mgh file overlaid onto lh of fsaverage_sym in tksurfer). Oddly, the vertex by vertex diff values also appear to be lower than we would expect (approximately by a factor of 10). Within the terminal output from glm_fit, we see both F and sig values of 0.
We are unsure as to why the analysis isn't working quite as expected. Any advice on this would be greatly appreciated.
Many thanks in advance,
Dan
---
# register subs
foreach sub (01 03 04 05 06 08 10 11 12 13 14 15 16 18 19 20 21 22 23 24 25 27 28 29 30 31 32 33 35 36 37 38 39 40)
surfreg --s S$sub-synth --t fsaverage_sym --lh
surfreg --s S$sub-synth --t fsaverage_sym --lh --xhemi
end
# mris_preproc
mris_preproc --target fsaverage_sym --hemi lh --xhemi --paired-diff --srcsurfreg fsaverage_sym.sphere.reg --meas curv --out lh.lh-rh.curv.sm00.mgh \
--s S01-synth --s S03-synth --s S04-synth --s S05-synth --s S06-synth --s S08-synth --s S10-synth --s S11-synth --s S12-synth --s S13-synth --s S14-synth \
--s S15-synth --s S16-synth --s S18-synth --s S19-synth --s S20-synth --s S21-synth --s S22-synth --s S23-synth --s S24-synth --s S25-synth --s S27-synth \
--s S28-synth --s S29-synth --s S30-synth --s S31-synth --s S32-synth --s S33-synth --s S35-synth --s S36-synth --s S37-synth --s S38-synth --s S39-synth \
--s S40-synth
# smooth
mris_fwhm --s fsaverage_sym --hemi lh --cortex --smooth-only --fwhm 5 --i lh.lh-rh.curv.sm00.mgh --o lh.lh-rh.curv.sm05.mgh
# mris_glmfit
mri_glmfit --y lh.lh-rh.curv.sm05.mgh --glmdir glm.lh.lh-rh.curv.sm05.osgm --osgm --surf fsaverage_sym lh
---
Daniel Carey
Marie Curie Ph.D. Fellow,
Centre for Brain and Cognitive Development,
32 Torrington Sq.,
London WC1E 7HX