FIX-Denoised Resting State fMRI data, 3T vs. 7T

[Ingrid Framås Syversen]

Hello,

I am using the WU-Minn HCP 1200 Subjects Data, and I am going to analyze the subset of 184 subjects with 7T data. For this, I have downloaded the "Resting State fMRI FIX-Denoised (Extended)" packages. However, when looking at the 3T and 7T timeseries, they appear different - the 7T timeseries looks like it is de-meaned, while the 3T timeseries is not (similar to what is discussed in this thread: https://groups.google.com/a/humanconnectome.org/g/hcp-users/c/eXYQuROqzUU/m/BSsyUIO5FAAJ). From this, it seems like the 3T data only underwent single-run FIX, while the 7T data underwent multi-run FIX. Is this correct? And is this described anywhere (I could not find it in the Reference Manual or other documents). Furthermore, if this is actually the case - is there anywhere I can access identically preprocessed + FIX-Denoised 3T and 7T rfMRI data, or do I have to do more denoising myself (if so, how)?

Best regards,

Ingrid Framås Syversen

[Glasser, Matthew]

We recommend you use data named ${StudyFolder}/${Subject}/MNINonLinear/${Results}/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean.dtseries.nii.  Before combining across runs, the data would ideally be demeaned and variance normalized.  The only released HCP data that was run through multi-run FIX is the 7T retinotopic data.  The resting state data was run through single run FIX.

 

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/f18e65ee-317e-4fa4-9d75-93b4ef4f78fdn%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.

[Ingrid Framås Syversen]

Hi,

I should probably have mentioned: I am looking at the files named ${fMRIName}hp2000_clean.nii.gz (I would like to use nifti files for my analyses and not cifti). Should these not be the same data as the ones you recommended, only with different file formats?

Best regards,

Ingrid Framås Syversen

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/zQMkuR-xcgY/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/C657234D-58DB-4784-92C1-30D670DE4684%40wustl.edu.

[Ingrid Framås Syversen]

Hi again,

I did not look at the cifti files before as I am not planning on using them, but I tried to directly compare the cifti and nifti files now. As far as I understand, ${fMRIName}_Atlas_1.6_hp2000_clean.dtseries.nii and ${fMRIName}hp2000_clean.nii.gz should be the same data, but in different formats. For the 3T data (two upper screenshots), it looks OK. However, in the 7T case (two bottom screenshots), the nifti files look "de-meaned", because there are both positive and negative voxel values, while the cifti files only have positive voxel values. This is why I am asking if the data were preprocessed the same way.

Best regards,

Ingrid Framås Syversen

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

It is not simply a different format, the cifti data has the cortex mapped onto surface vertices, which means it can be compared more accurately across subjects, surface-based smoothing won't pull in noise from white matter, or signal from opposing banks of sulci, etc.  Additionally, structures that aren't gray matter are not included in the cifti file.  There is also a small amount of smoothing applied to the cifti version.

Tim

---

From: Ingrid Framås Syversen <ingri...@gmail.com>

Sent: Thursday, March 4, 2021 10:57 AM
To: HCP-Users <hcp-...@humanconnectome.org>
Cc: Ingrid Framås Syversen <ingri...@gmail.com>
Subject: Re: [hcp-users] FIX-Denoised Resting State fMRI data, 3T vs. 7T

 

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/9fcc734d-1874-4035-8a93-816735e5a606n%40humanconnectome.org.

[Harms, Michael]

 

As you’ve observed, the 7T and 3T data were not processed identically as regards the application of “single-run” FIX (applied to REST scans for 3T, and REST and MOVIE scans for 7T).

 

The absence of the mean is due to a change in the behavior of ‘fslmaths -bptf’ (starting with FSL 5.0.7), such that it no longer included the temporal mean in its output.  This is relevant because we used FSL 5.0.6 for the processing of the 3T data, but FSL 5.0.9 for the processing of the 7T data.  I believe that we consequently changed the HCPpipelines code to account for this (i.e., to store the mean and add it back in), but obviously that wasn’t in place for the processing of the 7T data through single-run FIX.

 

That said, for something like computing connectivity, the mean is irrelevant.

 

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: Ingrid Framås Syversen <ingri...@gmail.com>

Reply-To: "hcp-...@humanconnectome.org" <hcp-...@humanconnectome.org>
Date: Thursday, March 4, 2021 at 10:57 AM
To: HCP-Users <hcp-...@humanconnectome.org>
Cc: Ingrid Framås Syversen <ingri...@gmail.com>

Subject: Re: [hcp-users] FIX-Denoised Resting State fMRI data, 3T vs. 7T

 

* External Email - Caution *

Hi again,

 

I did not look at the cifti files before as I am not planning on using them, but I tried to directly compare the cifti and nifti files now. As far as I understand, ${fMRIName}_Atlas_1.6_hp2000_clean.dtseries.nii and ${fMRIName}hp2000_clean.nii.gz should be the same data, but in different formats. For the 3T data (two upper screenshots), it looks OK. However, in the 7T case (two bottom screenshots), the nifti files look "de-meaned", because there are both positive and negative voxel values, while the cifti files only have positive voxel values. This is why I am asking if the data were preprocessed the same way.

 

Best regards,

Ingrid Framås Syversen

 

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/9fcc734d-1874-4035-8a93-816735e5a606n%40humanconnectome.org.

[Glasser, Matthew]

The mean can be obtained from uncorrected data if necessary.

 

Matt.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/3866054C-CAAC-4D8B-8635-10172660A036%40wustl.edu.

[Yashasvi Baweja]

Hello Ingrid and HCP maintainers,

Thank you for posting this discussion. I had a doubt for downloading resting state fMRI 7T denoised data, particularly, where were you able to locate the {$rfMRIName}_hp2000_clean.nii.gz data on s3 or db.hummanconnectome.org(curl+REST option)? Because when I try to look up for rfMRI_REST1_PA(particularly the ones acquired using 7T) under MNINonLinear/Results/ only rfMRI_REST1_(LR/RL) are listed (which I believe are 3T data). I am trying to download the data on a remote system with no GUI and it seems 7T rfMRI denoised data is only available via web browser. Apologies for posting off topic query, but if you could provide some assistance on how to get data via command line options, that'd really helpful!

Thanks,

Yashasvi Baweja

[B Demiral]

Hi Yashasvi,

I am not sure you had any response to your e-mail, but I just had an issue and was searching the mailing list.

I am using python pyxnat interference and use commands (somethings like below) to download purely from internet. 

For some subjects it can download the hp_2000_clean datasets from the FIX package under 7T, but for some others it cannot.

I can download unprocessed files for all of these subjects. I can also download some asc files under unprocessed package.

Luckily I have the hard disk copies where these files exist. But I was wondering why xnat does not let us to download them through internet.

Any ideas?

Baris

import pyxnat

cbd = pyxnat.Interface('https://db.humanconnectome.org',XNAT_USER,XNAT_PASSWORD)

hcp1200 = cbd.select.project('HCP_1200')

run_id='rfMRI_REST'

run_ap='AP'

pkg = run + '_FIX'
xnat_fixed = hcp1200.subject(str(sbj)).experiment(str(sbj)+'_7T').resource(pkg).file('rfMRI_'+run_id+'_7T_'+run_ap+'/rfMRI_'+run_id+'_7T_'+run_ap+'_hp2000_clean.nii.gz')