Pulling HCP Derivatives into CONN Toolbox

[Detroit Brain User]

Hello, 

I am attempting to extract the BOLD functional image (normalized to the MNI space) from the Minimally Processed HCP Pipeline (development ages 5-21), for further denoising and group-level preprocessing in CONN Toolbox. I would like to get a bit more information on how which file within the derivative folders this is. 

Thank you!

[Glasser, Matt]

We wouldn’t recommend making comparisons across subjects in MNI space.  It isn’t very accurate in the cerebral cortex (https://www.pnas.org/content/115/27/E6356.short).  The recommended data to use is in CIFTI space and has been functionally aligned across subjects and cleaned of spatially specific noise:

${StudyFolder}/${Subject}/MNINonLinear/Results/rfMRI_REST/rfMRI_REST_Atlas_MSMAll_hp0_clean.dtseries.nii.

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/46ac3c45-5a94-4787-a9e3-8aea57d3ec4fn%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.

[Mariam Reda]

Okay, that is helpful. I am looking through the directory, and this is the output.  Is that the file highlighted in green? Is there documentation for conducting band-pass filtering and seed-based analysis with this file? 

Also, would I be able to use the data file highlighted in yellow for denoising, first-level and second-level processing in an alternate program such as CONN? I.e., is that a CIFTI file as well?

/subject#/ses-None/files/MNINonLinear/Results/task-rest01

brainmask_fs.2.0.nii.gz         task-rest01_AtlasSubcortical_s2.nii.gz

DCANBOLDProc_v4.0.0             task-rest01_Jacobian.nii.gz

Movement_AbsoluteRMS_mean.txt   task-rest01.L.atlasroi.32k_fs_LR.func.gii

Movement_AbsoluteRMS.txt        task-rest01.L.native.func.gii

Movement_Regressors_dt.txt      task-rest01.nii

Movement_Regressors.txt         task-rest01.R.atlasroi.32k_fs_LR.func.gii

Movement_RelativeRMS_mean.txt   task-rest01.R.native.func.gii

Movement_RelativeRMS.txt        task-rest01_s2.atlasroi.L.32k_fs_LR.func.gii

RibbonVolumeToSurfaceMapping    task-rest01_s2.atlasroi.R.32k_fs_LR.func.gii

task-rest01_Atlas.dtseries.nii  task-rest01_SBRef.nii.gz

Thank you very much!

You received this message because you are subscribed to a topic in the Google Groups "HCP-Users" group.
To unsubscribe from this topic, visit https://groups.google.com/a/humanconnectome.org/d/topic/hcp-users/KhJlixM1EAQ/unsubscribe.
To unsubscribe from this group and all its topics, 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/F245204E-DFCE-45E0-9C3C-B6072BA73CBB%40wustl.edu.

--

Mariam Reda 

Co-​Founder & Research Director | Heal-Move-Shift

Research Assistant | ​Psychiatry Affective Neuroimaging Lab (PAN Lab)

B.S. Biopsychology, Cognition and Neuroscience & B.S. Creative Writing

College of LS&A | University of Michigan Class of 2019

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

Yes, the .dtseries.nii is a timeseries cifti file, but we would recommend the MR FIX cleaned version instead.  wb_command -cifti-average-roi-correlation may be a starting point for seed-based analysis, though we would also suggest parcellation-based analysis (wb_command -cifti-parcellate and then -cifti-correlate or similar), to get more signal while respecting the precise boundaries between functional areas (when using a good functional parcellation like the HCP MMP).  You can also read cifti files into python or matlab and go from there.

Uncompressed .nii without a cifti extension is not something I would expect to see in our data, I think someone on your end may have uncompressed it, and it is probably a volume timeseries (ordinary nifti-1).  We would not recommend using it instead of the MR FIX cleaned cifti files, due to volume registration's cross-subject inaccuracy in most of the cerebral cortex.

Tim

---

From: Mariam Reda <mhr...@umich.edu>
Sent: Tuesday, August 31, 2021 3:12 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: Re: [hcp-users] Pulling HCP Derivatives into CONN Toolbox

 

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/CAJ5g80Gg59dF_rhNm34pBSztcw3b6Hd9MKCV-oq2K7PJvCnJYQ%40mail.gmail.com.

[Glasser, Matt]

I am confused.  Are these your data or HCP Lifespan data?  If they are your data, you still need to run MR+FIX and MSMAll (including DedriftAndResample) Pipelines before doing any subsequent task or resting state analysis to ensure the data are cleaned and functionally aligned across subjects. 

 

We don’t typically do band pass filtering.  This is a very blunt approach that treats signal and noise equivalently. 

 

To make a parcellated connectome on a cleaned and functionally fMRI data, you first parcellate the data with wb_command -cifti-parcellate and then can compute the correlations with wb_command -cifti-correlation (it can also do covariance).  You can do more advanced network modeling with FSLNets (e.g. partial correlation).  To do group level analyses you can simply do things like a t-test across connectomes or linear regressions or that sort of thing.  The PALM software can also help with advanced statistics.  

 

Matt.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/SN6PR01MB4991127CC9F1AF1C36894F9581CC9%40SN6PR01MB4991.prod.exchangelabs.com.

[Harms, Michael]

 

Was the output even generated with the HCPPipelines?  The fact that there is a a “DCANBOLDProc” directory suggests that it was probably generated from the DCAN fork of the HCPPipelines.

 

Cheers,

-MH

 

-- 

Michael Harms, Ph.D.

-----------------------------------------------------------

Associate Professor of Psychiatry

Washington University School of Medicine

Department of Psychiatry, Box 8134

660 South Euclid Ave.                        Tel: 314-747-6173

St. Louis, MO  63110                          Email: mha...@wustl.edu

 

From: "Glasser, Matt" <glas...@wustl.edu>
Reply-To: "hcp-...@humanconnectome.org" <hcp-...@humanconnectome.org>
Date: Tuesday, August 31, 2021 at 7:26 PM
To: "hcp-...@humanconnectome.org" <hcp-...@humanconnectome.org>
Subject: Re: [hcp-users] Pulling HCP Derivatives into CONN Toolbox

 

* External Email - Caution *

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/D8916E84-9F81-49AE-9A93-19074328CE18%40wustl.edu.

[Mariam Reda]

Hello, 

Yes, I used the DCAN fork of the HCPPipelines in order to preprocess our own pediatric resting state data. 

Thank you all for your detailed feedback! I am looking to learn more about the commands you listed in order to begin analysis. I will also run the MR+FIX and MSMAll (including DedriftAndResample) Pipelines on the data before doing so.

Best,

Mariam 

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/71E0F535-BDB8-4C48-A756-4A48C4916341%40wustl.edu.

[Glasser, Matt]

Just to be clear—we have not affiliation with this DCAN fork so I don’t know if it keeps up with the upstream changes in the HCP Pipelines or if it diverged at some point.  I personally could not recommend using any forks that do not keep merging in upstream changes as we make bugfixes and improvements all the time to the HCP Pipelines.

Matt.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAJ5g80H4-JfO2N%2Bw%3DtZC%3DGhx5%2BDEXwyJXcMTQMkoitVyrqw5hA%40mail.gmail.com.

[Harms, Michael]

 

Adding to that:

I don’t know if output generated by the DCAN fork is fully compatible with the current versions of MR+FIX and MSMAll in the HCPpipelines.  It’s possible that the DCAN fork made changes which will either break those downstream pipelines or that the current version of MR+FIX and MSMAll in the HCPpipeline expect certain changes in other pipeline output that wasn’t part of the DCAN fork.  Last I checked, the DCAN fork has not been attempting to merge in the ongoing changes/improvements to the HCPpipelines, and it would probably be almost impossible to do so at this point in time.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/C7B3901B-837D-4300-A554-1F1442E04D6E%40wustl.edu.