Best way to align three mis-matched epi task runs when brain is cut off in some

AFNI Message Board
1

Hi folks,

I’ve been tasked with analyzing some older data that were never written up, that involve a scan procedure wherein for physiological monitoring and other procedures, the participant was removed from the scanner then re-entered and re-localized THREE times amid the single scan session. The three task runs were distributed in two (and sometimes three(!) different localizations, as was the MP-RAGE anatomical. Moreover, the scan montage did not cover the whole brain in some individual runs.

I have tried a whole different set of align_anat_epi.py and 3dAllineate scripts with different parameters, including -big_move, and including @align_centers as a preparatory step. However, it’s not working out. I’m not getting three aligned task runs, either with themselves, or with an anatomical base.

Could someone response with a suggested script language to cover this, especially when one task run EPI might have the top or the bottom of the brain cut off due to head positioning at the extremes of the collection space?

Thanks

Jim

2

It’s hard to say if some options like one of the “-partial_xxx” options might work. Maybe it would be better to upload some data, and I’ll take a look. I will PM you with instructions.

3

The problem was a bit odd here; the structural scan was larger than usual, and the centering with ginormous_move wasn’t enough to help here. I used the -cm option with @Align_Centers to align the center to the skullstripped dataset to start.

3dSkullStrip -prefix anat_ns -input structural_wholehead+orig.
@Align_Centers -base anat_ns+orig. -dset PSAPrun1vr+orig. -cm
@Align_Centers -base anat_ns+orig. -dset PSAPrun2vr+orig. -cm
@Align_Centers -base anat_ns+orig. -dset PSAPrun3vr+orig. -cm

Then “-giant_move” and “-dset1_strip None” to skip skullstriping with align_epi_anat.py seems to work well with all three EPI datasets.

4

Thanks so much Daniel!