x, y, z coordinates in XML file labels in custom FSL atlases

152 views
Skip to first unread message

Creence lin

unread,
Mar 27, 2024, 11:29:47 PM3/27/24
to hcp-...@humanconnectome.org

I am not sure how to create/find the x, y, z positions for XML files in the volumetric Glasser atlas as described in the 2016 Nature Paper A Multi-modal Parcellation of Human Cerebral Cortex https://www.nature.com/articles/nature18933 and the Hammersmith atlas described in the April 2017 Macroanatomy and 3D probabilistic atlas of the human insula paper https://www.sciencedirect.com/science/article/pii/S1053811917300964?via%3Dihub#s0160
The human insula is implicated in numerous functions. More and more neuroimaging studies focus on this region, however no atlas offers a complete subd…


For the Glasser atlas:
I was able to find the following data https://balsa.wustl.edu/study/RVVG, but it seems like you need to create an XML file with x, y, and z indices for each label and I am not sure if there was one already created or what those labels would be or which of the nii.gz files corresponds to the Glasser atlas.

So I used the data in the HCP-MMP1 link in https://neuroimaging-core-docs.readthedocs.io/en/latest/pages/atlases.html which has an XML file, but when I plot a point in FSL, they don't seem to correspond to a point on the brain picture.  I also used this volumetric data from https://figshare.com/articles/dataset/HCP-MMP1_0_projected_on_MNI2009a_GM_volumetric_in_NIfTI_format/3501911?file=5534027 , but there is no XML file so I just used the points in the XML file from thehttps://neuroimaging-core-docs.readthedocs.io/en/latest/pages/atlases.html link.

For the Hammersmith atlas, the paper had a link to https://soundray.org/hammers-n30r95/ and http://www.brain-development.org which if you go to Atlases -> Adult Brain Atlases -> Hammers Atlas Database Adult atlases (30 subjects, 95 regions), but when I download all the data, I never see the x, y, and z coordinate positions for the 95 regions in any of the files.
Rolf Heckemann's Website :: Professor of Medical Imaging and Image Analysis :: Medical imaging, neuroimaging, image processing, morphometric analysis
Any advice on how to do this or where to get the XML files for FSL?

Glasser, Matt

unread,
Mar 27, 2024, 11:43:02 PM3/27/24
to hcp-...@humanconnectome.org

We didn’t release a cortical parcellation in the volume.


Matt.

 

From: Creence lin <cre...@stanford.edu>
Reply-To: "hcp-...@humanconnectome.org" <hcp-...@humanconnectome.org>
Date: Wednesday, March 27, 2024 at 7:50 PM
To: "hcp-...@humanconnectome.org" <hcp-...@humanconnectome.org>
Subject: [hcp-users] x, y, z coordinates in XML file labels in custom FSL atlases

 

 

I am not sure how to create/find the x, y, z positions for XML files in the volumetric Glasser atlas as described in the 2016 Nature Paper A Multi-modal Parcellation of Human Cerebral Cortex https://www.nature.com/articles/nature18933 and the Hammersmith atlas described in the April 2017 Macroanatomy and 3D probabilistic atlas of the human insula paper https://www.sciencedirect.com/science/article/pii/S1053811917300964?via%3Dihub#s0160

Image removed by sender.

The human insula is implicated in numerous functions. More and more neuroimaging studies focus on this region, however no atlas offers a complete subd…

 

 

For the Glasser atlas:

I was able to find the following data https://balsa.wustl.edu/study/RVVG, but it seems like you need to create an XML file with x, y, and z indices for each label and I am not sure if there was one already created or what those labels would be or which of the nii.gz files corresponds to the Glasser atlas.

 

So I used the data in the HCP-MMP1 link in https://neuroimaging-core-docs.readthedocs.io/en/latest/pages/atlases.html which has an XML file, but when I plot a point in FSL, they don't seem to correspond to a point on the brain picture.  I also used this volumetric data from https://figshare.com/articles/dataset/HCP-MMP1_0_projected_on_MNI2009a_GM_volumetric_in_NIfTI_format/3501911?file=5534027 , but there is no XML file so I just used the points in the XML file from thehttps://neuroimaging-core-docs.readthedocs.io/en/latest/pages/atlases.html link.

 

For the Hammersmith atlas, the paper had a link to https://soundray.org/hammers-n30r95/ and http://www.brain-development.org which if you go to Atlases -> Adult Brain Atlases -> Hammers Atlas Database Adult atlases (30 subjects, 95 regions), but when I download all the data, I never see the x, y, and z coordinate positions for the 95 regions in any of the files.

Rolf Heckemann's Website :: Professor of Medical Imaging and Image Analysis :: Medical imaging, neuroimaging, image processing, morphometric analysis

Any advice on how to do this or where to get the XML files for FSL?

--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CH0PR02MB8026474C52C9C5DA1B5DBC5BCE342%40CH0PR02MB8026.namprd02.prod.outlook.com.

 


The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

Tim Coalson

unread,
Mar 27, 2024, 11:48:07 PM3/27/24
to hcp-...@humanconnectome.org
The Glasser 2016 HCP MMP 1.0 is not a volumetric atlas.  Another paper of ours explains why it isn't, and why any binary volumetric form of it should not be considered a faithful representation of the entire cortex:


However, the insula is a subset of cortex that has relatively low cross-subject variability in human subjects, so a version restricted to that region may be reasonable to represent as a volume file, if necessary.  We do have a probabilistic volume version of the entire MMP 1.0 in MNINonLinear space (mainly for the purpose of showing how blurry volumetric methods make it), but we didn't compute the centers of gravity, because they aren't a faithful representation of a non-spherical parcel, and we didn't expect the centers of gravity to be a required part of any format:


This version also takes partial-voxel membership into account:


Neither of these includes smoothing effects, so data that has been volumetrically smoothed will implicitly have substantially larger blurring and conflation than these probabilistic versions.

Tim


--

Creence lin

unread,
Mar 28, 2024, 1:25:20 PM3/28/24
to hcp-...@humanconnectome.org
Thanks for the clarification, Tim and Matt. Do you actually need to compute the centers of gravity to load the atlases into FSL?  I see so many atlases that just have the nii.gz file and no corresponding XML file that you would need to load the atlas into FSL.  If there are no centers of gravity how do people view these atlases in FSL?
Creence

From: Tim Coalson <tim.c...@gmail.com>
Sent: Wednesday, March 27, 2024 8:47 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: Re: [hcp-users] x, y, z coordinates in XML file labels in custom FSL atlases
 

Tim Coalson

unread,
Mar 28, 2024, 8:52:43 PM3/28/24
to hcp-...@humanconnectome.org
I am not particularly familiar with using atlases in FSL.  You might ask the FSL mailing list:


However, in accordance with our paper, if you are interested in human cortex outside of the insula or central sulcus, we recommend doing surface-based comparisons with our atlas in its original form, instead of figuring out how to do blurry volume-based comparisons.

Tim


Reply all
Reply to author
Forward
0 new messages