Nifti Coordinates and Atlases

1

I am looking for map the nifti coordinates to Atlas areas. I have below questions:

  • First I think my nifti coordinates are in i,j,k space[since both qform_code and sfrom_code == ‘scanner’], right? so I need to convert them to x,y,z[MNI space first].Can I just apply the affine transformation and it will give me the coordinates in MNI space or should I use the function http://nilearn.github.io/modules/generated/nilearn.image.coord_transform.html#nilearn.image.coord_transform? What confuses me is that the MNI coordinates I get are floats but I think they must be integers. The affine transformation in my nifit is:
    srow_x : [ -2.95961881 -0.204623 0.59447175 82.84215546]
    srow_y : [ -0.22385383 2.98946524 -0.15194418 -97.11065674]
    srow_z : [ 0.43651536 0.1456929 3.95265937 -88.87265015]

  • How to map each MNI coordinates[x,y,z] to atlas areas. which tool I can use for this, lets say If I want to map MNI space to Harvard-Oxford Atlas? I think FSL can do. Did anyone did something like this before? I also see http://nilearn.github.io/modules/reference.html#module-nilearn.regions has some functions, but I am not sure which one can be better to use/ or if nilearn can really be used for this purpose or not.

  • Can we say that Talairach and Brodmann Areas are similar?If I map to Talairach I also map to brodmann areas?

  • Another tool which seems like might work is: https://brainder.org/2012/07/30/automatic-atlas-queries-in-fsl/ but I do not understand what does threshold mean here.

2

About your first question, as you have qform_code and sfrom_code == ‘scanner’, it means that after applying “image.coord_transform(50, 50, 50, niimg.affine)” you will get the coordinates in the scanner coordination, and not MNI. Then if you want the MNI, you should convert these scanner reference to MNI coordinates.

3

About your third question, if you execute the snippet below, with each of the 5 levels (‘hemisphere’, ‘lobe’, ‘gyrus’, ‘tissue’, ‘ba’), you can see different segmentations in 1 constant coordination system, called Talairach.

Retreive the atlas

from nilearn import datasets

level_name : {‘hemisphere’, ‘lobe’, ‘gyrus’, ‘tissue’, ‘ba’}

Talairach_atlas = datasets.fetch_atlas_talairach(‘ba’)
atlas_filename = Talairach_atlas.maps
labels = Talairach_atlas.labels # 71 Brodmann areas
from nilearn import plotting
plotting.plot_roi(atlas_filename)