I am trying to locate the atlas datasets, i.e. TT_Daemon.Brodmannarea_*.BRIK/HEAD, for Brodmann areas 3 Left and Right to create masks.
I was told that they must be stored somewhere in the same directory as AFNI, but I couldn’t find them.
Could anyone please help. I am new to AFNI.
The Brodmann areas should be the second volume (sub-brick 1) of the TTatlas+tlrc dataset.
Thank you for your response.
So I am wondering what command should I use to extract the BRIK/HEAD files for each of the Brodmann areas 3 Left and 3 Right.
Sorry I am new to AFNI, and from this link ‘https://sscc.nimh.nih.gov/afni/doc/misc/afni_ttatlas/’, I found this line:
1 (sub-brick#) 83 (Value) Brodmann area 3
No separate values for the Left and Right areas.
The atlas codes are available in the atlas dataset itself
3dinfo -atlas_points TTatlas+tlrc
You can extract regions with the whereami command or any other afni command. See the examples in afni11_roi.pdf like this one:
whereami -mask_atlas_region TT_Daemon:left:hippocampus -prefix Lhip
I am interested in using whereami to generate anatomical masks in MNI space. Although the whereami -show_atlas_code indicates that the TT atlas has many regions from which I could generate a mask in tlrc space, I’m not seeing anything for the MNI_avg152T1 atlas I have using to co-localize subject structurals and activations. I also tried the 3dinfo command on this atlas.
I did whereami -show_atlas_code | less and see several other atlases with parcellations. Is it that there are several atlasas by different names in there that all rely on the spatial warping of MNI(152), but parcellate structures differently?
What should I do with the whereami -mask_atlas_region XX-xxx:yyyy etc? what atlas should I invoke?
The MNI_avg dataset is a template, not an “atlas” in our definition because it doesn’t have a segmentation/regions labeled within it. You can see the list of atlases available in your afni binary distribution by using
Also use the afni GUI to explore the atlases with
replacing ~/abin with wherever your datasets are stored.
I guess my question becomes which of the atlases that are in AFNI were spatially-oriented in MNI space, when making the spatial parcellations between brain structures? Is there a wiki or something on where each of these AFNI-furnished atlases came from and related source papers that would have this information?
To be sure, I can just use whereami to generate masks and overlay them on MNI structurals of my subjects and see if they make sense visually, it would just be nice to have confirmation that an atlas in AFNI was created in MNI space.
The whereami -show_atlases option shows the “space” of the datasets as the second column. For any dataset including atlases, you can also view the space with “3dinfo -space”.
I obtained the screen output that indicates the rough story on each atlas.
I was able to dump out amygdala masks for CA_ML_18_MNIA, however they looked too superior and lateral. Maybe it was just the sample subject’s underlay…
However, what would be the syntax to use one of the CAxxx MNI-based atlases that do not have a left or right specifier?
I got an error when I entered either:
whereami -mask_atlas_region CA_N27_PM:u:Amyg:6 (too many colons) or
whereami -mask_atlas_region CA_N27_PM:u:Amyg or
whereami -mask_atlas_region CA_N27_PM:Amyg:6 or
Don’t use :u: in the middle. The middle part is for left or right only. The atlases that have MNIA in their names are in MNI_ANAT space with the AC at (0,0,0). MNI space atlases are available here, or you can shift the MNI_ANAT atlases with 3dWarp.