Why register fMRI to topup fieldmap, rather than the other way around?

[Keith Jamison]

I'm referencing portions of the HCP pipeline for some non-HCP data, and noticed that the way fMRI is processed via the TopupPreprocessingAll.sh script is:

1. compute topup fieldmap from SE pair, creating an in-plane warp file

2. rigid-body align fMRI scout to SE scan

3. create distortion correction warp by concatenating (EPI->SpinEcho-space) + (SpinEcho-space topup unwarp)

Ideally, the SE pair and fMRI are well aligned at acquisition, but simple motion could lead to the fMRI being rotated such that the slices aren't perfectly aligned with the SE.

Would it make more sense to do something like:

1. compute topup fieldmap from SE pair

2. rigid-body align fMRI scout to SE scan

3. invert this transform so we have SE->fMRI 

3. resample the SE FieldMap (rad/s output from topup) to fMRI-space using SE->fMRI transform

4. create new in-plane warp that is now explicitly in fMRI slice space

Not suggesting a change to the HCP pipelines really, but curious about the logic for this step in general.

Thanks!

-Keith

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

I suspect the thinking at the time was to avoid the extra complication of generating a second warpfield to keep track of, and extra steps.

Given the tools available, currently we can only immediately generate a distortion-corrected version of the SE scans, which means the fMRI scout has to initially be aligned while it is still distorted.  If there is motion such that the distortions occur at a different angle, these will still interfere with a standard registration that expects the same distortion axis as the SE images, leaving some possible alignment error anyway.  There is some ongoing work on improving distortion correction.

Tim

---

From: Keith Jamison <keith....@gmail.com>
Sent: Friday, November 19, 2021 6:06 PM
To: HCP-Users <hcp-...@humanconnectome.org>
Subject: [hcp-users] Why register fMRI to topup fieldmap, rather than the other way around?

 

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.

--
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/52120940-b35e-4c07-a00e-6cab1e9f43ddn%40humanconnectome.org.

[Keith Jamison]

That's a fair point. I haven't played with it, but I know FLIRT accepts a fieldmap argument that it applies during registration, thus allowing a fit between distorted and undistorted images.

-Keith

[Harms, Michael]

 

The Oxford folks are working on a tool that would enable correction for the interaction between susceptibility and motion (similar to what is available in ‘eddy’ for dMRI).

 

-- 

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: Keith Jamison <keith....@gmail.com>
Reply-To: "hcp-...@humanconnectome.org" <hcp-...@humanconnectome.org>
Date: Friday, November 19, 2021 at 8:29 PM
To: HCP-Users <hcp-...@humanconnectome.org>
Cc: "tsc...@mst.edu" <tsc...@mst.edu>
Subject: Re: [hcp-users] Why register fMRI to topup fieldmap, rather than the other way around?

 

* External Email - Caution *

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/17b912c6-bdc8-4ccb-bfc1-512cc607885en%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.

[Keith Jamison]

Glad to hear it. Can't wait to see what clever new acronym they come up with!

Keith

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/728AF6DB-24AA-4A0F-B466-B017A7025971%40wustl.edu.