Fmriprep and native space

, ,
1

Hello,

I’m trying to do an analysis in native space and want to better understand how fmriprep calculates native space to make sure I’m aligning my files correctly. The plan would be to use freesurfer segmentation for creating native masks of my ROIs.

fslhd '/.../bids/derivatives/fmriprep-20.2.7/fmriprep/sub-011220701/**anat**/sub-011220701_desc-aparcaseg_dseg.nii.gz' 

|qto_xyz:1|1.000000 0.000000 0.000000 -100.189316 |
|---|---|
|qto_xyz:2|0.000000 1.000000 0.000000 -102.280685 |
|qto_xyz:3|0.000000 0.000000 1.000000 -119.794266 |
|qto_xyz:4|0.000000 0.000000 0.000000 1.000000 |

fslhd '/.../bids/derivatives/fmriprep-20.2.7/fmriprep/sub-011220701/**func**/sub-011220701_task-trivia_run-1_space-T1w_desc-brain_mask.nii.gz' 

qto_xyz:1	1.200000 0.000000 0.000000 -70.189316 
qto_xyz:2	0.000000 1.200000 0.000000 -55.480682 
qto_xyz:3	0.000000 0.000000 1.200000 -59.794262 
qto_xyz:4	0.000000 0.000000 0.000000 1.000000

Shouldn’t these two files have the same affine? Or can I just use nilearn.image.resample_to_img to align these two files?

2

Hi @foldes.andrei , yes it is exactly as you said, the two images are in the same space (they should overlap perfectly when you look at them in a viewer) but on a different grid. You would need to resample the dseg file to the bold grid if you want to apply the ROIs to the bold image.