Combining T1 & T2 scans - warped models with excess tissue outside of the head

42 views
Skip to first unread message

Jake E Toth

unread,
Feb 4, 2022, 12:22:56 PM2/4/22
to roast-users
Hello,

I am using both T1 and T2 MRI scans,  individually these appear normal when displayed with SPM. However, once they have been aligned and coregistered the models are often warped and produce elongated alien-like head models with excess material in the form of blocks of CSF that extend out of the neck.

The T2 scans are cropped (do not include the neck) and these crops are at different angles. The T1 scans do include the neck. I think the issue is with SPM, and is caused by these scans not having an identical field of view/cropping of the head. (Apologies if that is the wrong terminology for MRI scans, but I hope you can appreciate the point.)

Although not ideal I can trim the excess material manually, but the warping is a harder issue to solve. I have tried to manually set the origin in the scans to be identical which did not help.

Has anyone ran into this issue before and been able to solve it? Or is it recommended to only use T1 and T2 scans if the fields of view are the same so SPM can easily combine the scans?

Kind regards,
Jake 




Yu (Andy) Huang

unread,
Feb 5, 2022, 11:48:43 AM2/5/22
to Jake E Toth, roast-users
Hi Jake,

Based on my experience, SPM usually works great for aligning T1 and T2 images. If you try to align them in SPM using "coreg" or "normalise" and both fail, I suspect the images have bad headers. I would check the header info and fix the header first before running registration between T1 and T2.

Hope this helps.

--
You received this message because you are subscribed to the Google Groups "roast-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to roast-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/roast-users/e7a9a863-f5b1-424c-aa61-e157f4b5e49cn%40googlegroups.com.


--
Yu (Andy) Huang, Ph.D.

Research Associate, Radiology Dept., Memorial Sloan Kettering Cancer Center
Rm. 321-G17P, 321 East 61st Street, New York, NY 10065
Office: 646-608-7608

Research Associate, CCNY-MSK Partnership for Artificial Intelligence
Center for Discovery and Innovation, Rm. 3.320,
85 St Nicholas Terrace, New York, NY 10031
Email: yhua...@citymail.cuny.edu; andypo...@gmail.com
Web: http://www.parralab.org/people/yu-andy-huang/

Davide

unread,
Feb 8, 2022, 10:52:18 AM2/8/22
to roast-users
I had the same issue and I think it was related to artefacts in my T2 scans. 

Jake E Toth

unread,
Feb 8, 2022, 11:12:01 AM2/8/22
to roast-users
Interesting, what artefacts did you observe in the T2 scans?
The T2 scans I am using appear normal, but are cropped and rotated when compared to the T1 scans.
Message has been deleted

Davide

unread,
Feb 8, 2022, 11:17:32 AM2/8/22
to roast-users
In my scans it looked like there was some drop in the signal and I believe this was making it difficult for spm to align the T1 and T2 scans, maybe? My masks looked like yours, they had an alien-like shape.

Jake E Toth

unread,
Feb 8, 2022, 11:23:16 AM2/8/22
to roast-users
Thank you, I will keep that in mind and I will update this thread if I find a resolution to this problem.

Davide

unread,
Feb 11, 2022, 5:47:18 AM2/11/22
to roast-users
Do you have the resampling option turned on?

Jake E Toth

unread,
Feb 11, 2022, 7:21:31 AM2/11/22
to roast-users
Yes and I have tried with resampling off as well and it does not solve the problem, unfortunately.

Thanks
Reply all
Reply to author
Forward
0 new messages