Hi AFNI Experts,

I am trying to run a multiple comparison correction for an ANOVA.

I have generated an average group mask (using each volunteer's brain mask after being warped to standardized space) as follows:
3dMean –prefix avg_group_mask mask_sub1+orig mask_sub2+orig mask_sub3+orig…
→ Output: avg_group_mask+orig.{HEAD|BRIK}

I checked to see whether the mask had the same dimensions as my statistical dataset:
3dinfo –prefix –d3 –n4 ANOVA3_result_edit+tlrc avg_group_mask+orig
→ Output:
ANOVA3_result_edit -2.000000 -2.000000 2.000000 91 109 91 34
avg_group_mask -3.125000 3.125000 3.965232 64 64 40 1

Since the dimensions were different, I used 3dfractionize to adjust the dimensions:
3dfractionize –template ANOVA3_result_edit+tlrc –input avg_group_mask+orig –clip 0.2 -preserve -prefix avg_group_mask_corr
→ Output: avg_group_mask_corr+tlrc.{HEAD|BRIK}

I checked the dimensions again, and they looked good:
3dinfo -prefix -d3 -n4 ANOVA3_result_edit+tlrc avg_group_mask_corr+tlrc
→ Output:
ANOVA3_result_edit -2.000000 -2.000000 2.000000 91 109 91 34
avg_group_mask_corr -2.000000 -2.000000 2.000000 91 109 91 1

I re-ran my ANOVA3 script through the mask (having added -mask avg_group_mask_corr+tlrc).

When I looked at the results in the AFNI viewer, it appears that the results no longer overlay properly onto my averaged anatomical volume (ave_anatomy+tlrc).

NB. the average anatomical volume was created by:
@auto_tlrc –base TT_avg152T1+tlrc –input run1-T1+orig –suffix _ref_avg152T1 –ok_notice –OK_maxite
→ Output: run1-T1_ref_avg152T1+tlrc.{HEAD|BRIK}

@auto_tlrc –apar run1-T1_ref_avg152T1+tlrc –input run1-T1+orig –suffix _avg152T1 –OK_maxite
→ Output: run1-T1_avg152T1+tlrc.{HEAD|BRIK}

3dmerge –prefix ave_anatomy -1blur_fwhms 3 run1-T1_avg152T1... {added all volunteer’s structural scans}
→ Output: ave_anatomy+tlrc.{HEAD|BRIK}

I checked the avg_group_mask+orig (which was generated from 3dMean of each volunteer's brain mask using 3dAutomask), and it appears to overlay nicely with ave_anatomy+tlrc. However, the resampled average group mask (avg_group_mask_corr+tlrc) seems to have moved down and forward when overlaid on the anatomical volume. When looking at the ANOVA3 results after being run through this corrected mask, it appears to have also moved down and forward identically.

I am wondering what might have happened, and how to go about fixing the descrepancy? I am wondering if 3dfractionize might have caused the shift in the mask, and if I had perhaps used it incorrectly, and if so, how I can remedy the error?

I am very new to AFNI, so any help would be greatly appreciated (and if any other clarrification is required, please let me know!)

Thank you in advance.