ICA FIX for Young Adult tFMRI

125 views
Skip to first unread message

Micha Burkhardt

unread,
Jan 6, 2023, 7:50:35 AM1/6/23
to HCP-Users
Hi HCP team,

I'm planning on running dynamic functional connectivity analyses on tfMRI data and I was wondering what the recommended preprocessing steps are in that case.
  1. To my knowledge, the Young Adult tFMRI data is currently not processed by the ICA FIX pipeline? Is this correct?
  2. Is it possible (and recommended) to manually use the ICA FIX pipeline on the provided (minimally processed) tFMRI CIFTI data? 
  3. Is additional tICA also recommended (or not yet ready for users)?
  4. If this (or some of it) is possible, are there any other recommended processing steps before parcellation and time series extraction (like rerunning MSMAll)?
These questions seem quite trivial. However, I was unable to find any specific information for tfMRI data in this regard, and any help would be greatly appreciated. Thanks!

Best,
Micha

Glasser, Matt

unread,
Jan 6, 2023, 7:38:09 PM1/6/23
to hcp-...@humanconnectome.org

The released tfMRI data for HCP-YA was not processed with sICA+FIX.  3T and 7T resting state and task fMRI data for the 1071 subjects with MSMAll has been run through sICA+FIX, tICA cleanup, and has had the bias correction fixed.  The next step is packaging the data and releasing it.  We hope to do this soon, but as always there are many competing priorities.

 

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/7818b690-e0ac-4842-8cd4-964ae49a5976n%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.

Micha Burkhardt

unread,
Jan 9, 2023, 5:05:54 AM1/9/23
to HCP-Users, glas...@wustl.edu
Dear Matt,

Thanks for the quick reply! Is it feasible to do it with the current processing pipeline, or would it make more sense to wait for your data release?

Best,
Micha

Glasser, Matt

unread,
Jan 9, 2023, 7:03:27 AM1/9/23
to hcp-...@humanconnectome.org

Coles, Claire

unread,
Jan 9, 2023, 9:03:14 AM1/9/23
to hcp-...@humanconnectome.org, glas...@wustl.edu

Please unsubscribe me from this list serve

 

From: Micha Burkhardt <micha.b...@informatik.tu-chemnitz.de>
Sent: Monday, January 9, 2023 5:06 AM
To: HCP-Users <hcp-...@humanconnectome.org>

James Thompson

unread,
Apr 12, 2023, 4:14:14 PM4/12/23
to HCP-Users, glas...@wustl.edu
Hi Matt,
Is there any word on the release of the HCP-YA tfMRI data with sICA+FIX? I too am eager to use it. Appreciate the many competing concerns though!
Jim

Glasser, Matt

unread,
Apr 12, 2023, 8:43:29 PM4/12/23
to James Thompson, HCP-Users

Unfortunately, that is the key issue right now.  Sorry for the delay.


Matt.

Micha Burkhardt

unread,
Aug 8, 2023, 11:04:24 AM8/8/23
to HCP-Users
Hi Matt,

Sorry for another mail concerning this topic. Are there any updates for the release of the cleaned tfMRI data for HCP-YA?

And on a second note, what would you recommend as the most important additional steps (if any) for basic preprocessing if the cleaned data is not yet available? Pipelines in literature vary wildly, but I am sure you have some well informed opinions in that regard.

Many thanks,
Micha

Glasser, Matt

unread,
Aug 8, 2023, 11:20:57 AM8/8/23
to hcp-...@humanconnectome.org

Hoping to work on moving this forward at our next group meeting.  One issue is that we have been focusing on the HCP-Lifespan release, which will also have full spatial and temporal ICA cleanup. 

 

At this point, we are using linear detrending, spatial ICA, temporal ICA and Wishart Filtering/parcellation/nothing depending upon application. This means no spatial smoothing, temporal smoothing, tissue-based nuisance regression or derivative methods (e.g., compcor), global signal regression, motion parameter regression, physiological regressors, or scrubbing.

Reply all
Reply to author
Forward
0 new messages