A Question regarding rs-fMRI data

[najme soleimani]

Dear HCP developers,

I want to generate functional connectivity matrix using rs-fMRI data.

I have my own parcellation (based on Desikan killiany atlas and in different scales) which is generated with FreeSurfer. I wonder if I can use rfMRI_REST1_LR_hp200_clean.nii.gz as input?

on the other hand, for each subject I have 4 RUNs. Is it possible to calculate functional connectivity matrix for each run and then merge these matrix connectivities?

I wonder if I should dmean and normalize the rfMRI_REST1_LR_hp200_clean.nii.gz data before generating the functional connectivty matrix?

I would appreciate if you could help me...

Best regards,

Najme

[Glasser, Matt]

A folding-based parcellation is not going to respect functional boundaries.  I would use a functionally relevant parcellation like the HCP’s multi-modal parcellation:

https://balsa.wustl.edu/file/3VLx

 

You should use the MSMAll aligned cleaned resting state data:

${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean.dtseries.nii

 

You should demean and can variance normalize as described here:

https://wiki.humanconnectome.org/display/PublicData/HCP+Users+FAQ topic #3

 

Matt.

--
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/79ff917f-5eb2-4d1a-bebf-2c8e97730448n%40humanconnectome.org.

 

---

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.

[Coalson, Timothy Scott (S&T-Student)]

In particular, if you want to generate connectivity using multiple runs, you should concatenate the dtseries before correlation to avoid making multiple dconn files.  The demeaning and optional variance normalization before concatenation is just to prevent the concatenation from producing artifacts due to discontinuities.  If you plan to only use parcellated results, you should parcellate the dtseries before correlating to get better SNR.

If you plan to look at the dense correlation, you still may not need to make a dconn, as wb_view can correlate dtseries on the fly with its dynconn feature.

Tim

---

From: Glasser, Matt <glas...@wustl.edu>
Sent: Tuesday, November 9, 2021 9:57 AM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: Re: [hcp-users] A Question regarding rs-fMRI data

 

WARNING: This message has originated from an External Source. This may be a phishing expedition that can result in unauthorized access to our IT System. Please use proper judgment and caution when opening attachments, clicking links, or responding to this email.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/4086D421-F4ED-404C-9A5F-766FD9570D23%40wustl.edu.

[najme soleimani]

Hi,

Thanks for your quick reply...

I opened the cifti file that you mentioned (  ${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean.dtseries.nii) using connectome workbench (wb_view). I also used the wb_command  -cifti-separate and  separate all volume structures of this data into a volume in order to be able to check it with MRIcro. I could see the subcortical voxels but I could not see the cortical voxels. I wonder how I can see the cortical voxels...

and also which *.surf.gii is relevant to the (${fMRIName}_Atlas_MSMAll_hp2000_clean.dtseries.nii ) cortical coordinates?

kind regards,

Najme

[Joseph Orr]

The volume component of the cifti doesn't have cortical voxels, the cortex is represented as surfaces. You might find it helpful to look at 'wb_command -cifti-help' or these posts: https://mandymejia.com/2015/08/10/a-laymans-guide-to-working-with-cifti-files/

https://mvpa.blogspot.com/2014/03/nifti-cifti-gifti-in-hcp-and-workbench.html

--

Joseph M. Orr, Ph.D.

Assistant Professor

Department of Psychological and Brain Sciences

https://tamu.zoom.us/my/josephorr

On Wed, Nov 10, 2021 at 1:26 PM najme soleimani <najme.sol...@gmail.com> wrote:

Hi, Thanks for your quick reply... I opened the cifti file that you mentioned ( ${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean.dtseries.nii) using connectome workbench (wb_view). I also used ZjQcmQRYFpfptBannerStart

This Message Is From an External Sender

This message came from outside your organization.

ZjQcmQRYFpfptBannerEnd

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/3a905465-5f41-4473-84ca-0169d3603210n%40humanconnectome.org.

[Coalson, Timothy Scott (S&T-Student)]

FYI, for MRI data, we don't recommend using fieldtrip functions for using cifti in matlab like that first link does, as they make some changes to the data compared to how they are stored on disk (they pad the matrix to fill the medial wall with NaNs), and translates the header information into a structure used for MEG/EEG data.  The relatively new cifti-matlab v2 library gives you the file data as it exists on disk, and the header information as it is arranged in the cifti file:

https://github.com/Washington-University/cifti-matlab

Tim

---

From: Joseph Orr <josep...@tamu.edu>
Sent: Wednesday, November 10, 2021 2:16 PM
To: HCP Users <hcp-...@humanconnectome.org>

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAOxRjYny9o4eaNZFXg67%3D1QEqABw61EdF2FmKg_rG8%2BuGccswQ%40mail.gmail.com.