Dear Naj and Mark,

Thanks for your help. Sorry for the confusion, I had used the control image of the ASL sequence for flirt 6dof -nosearch registration to the T1w*. I then applied that registration to the CBF (generated through Functool on the GE scanner), just to demonstrate the overlay.

Naj, you mentioned the blue bands, "Also, those blue bands running across the brainstem into posterior corpus callosum are more alarming than prefrontal distortion."; even after inhomogeneity correction in SPM (separately for labelled and tagged images) and subsequent calculation of CBF in asl_file** these are not mitigated. Do you know what those blue bands are, is it a common artefact in ASL-derived data?

Overall, I am also really very confused as to why the CBF maps only have negative values (gray matter cbf around -150 and white matter around -500). Just taking the absolute value would result in low GM CBF and high WM CBF, which is not correct, so I am unsure how to even deal with that.

Thank you again for helping.
Best wishes,
Barbara

* flirt -in c002_controlASL_brain.nii.gz -ref c002_T1brain.nii.gz -omat c002.mat -out c002controlASL2T1brain.nii -dof 6 -nosearch
flirt -in c002_ASLbrain.nii.gz -ref c002_T1brain.nii.gz -init c002.mat -applyxfm -out c002_ASLbrain2T1.nii.gz

** asl_file --data=../c002_rawASL.nii.gz --ntis=1 --iaf=ct --diff --out=diffdata --mean=diffdata_mean



On Mon, Jul 27, 2015 at 7:48 AM, Mark Jenkinson <[log in to unmask]> wrote:
Hi,

I’m afraid that without fieldmaps or any phase-encode reversed acquisitions then there are no tools in FSL for distortion correction.

As for registrations - you should try and use an image containing anatomical information (e.g. not a CBF map) when driving the registrations.  Also, as Naj pointed out, non-linear registrations can behave badly on low resolution or noisy data and result in more distortions rather than less.  We have occasionally used non-linear registration to help with distortion correction, but the results are highly variable and require good data and very careful evaluation to make sure that it is not moving things in directions other than the phase-encode direction (as that is the only direction that distortions will occur in).   I would try using BBR with some data from your sequence that has the best grey/white-matter contrast, and see if that improves your registrations.

All the best,
Mark


From: FSL - FMRIB's Software Library <[log in to unmask]> on behalf of "Najmeh Khalili-Mahani, Dr" <[log in to unmask]>
Reply-To: FSL - FMRIB's Software Library <[log in to unmask]>
Date: Friday, 24 July 2015 16:30
To: "[log in to unmask]" <[log in to unmask]>
Subject: Re: [FSL] GE resting-state ASL

Hi Barbara,

So this is a subject CBF map registered onto a BETed subject T1W? 

If so, more than distortion (which is not so terrible) I would focus on improving BET (which appears to be removing too much from the prefrontal region). After that, you can improve alignment by nonlinear registration, or multi-resoluton iterative linear registration but this probably won't buy you much improvement in specificity (because of PVE). Also, those blue bands running across the brainstem into posterior corpus callosum are more alarming than prefrontal distortions. A nonlinear registration may introduce more errors. 

Hth.

Best
Naj



On Fri, Jul 24, 2015 at 7:05 AM, Barbara Kreilkamp <[log in to unmask]> wrote:
Sure,
Thank you. They are EPI-distortions in the phase encode direction (Y-axis), marked with green epilpses on the screenshot -> find the image here please: https://app.box.com/s/7noguefmvevhu44kso8pdh58dtsehdps
Unfortunately I do not have field maps.
Best wishes,
Barbara

On Fri, Jul 24, 2015 at 11:06 AM, Barbara Kreilkamp <[log in to unmask]> wrote:



Sure,
Thank you. They are EPI-distortions in the phase encode direction (Y-axis), marked with green epilpses on the screenshot.
Unfortunately I do not have field maps.
Best wishes,
Barbara


On Thu, Jul 23, 2015 at 8:50 PM, Najmeh Khalili-Mahani, Dr <[log in to unmask]> wrote:
Hi,

Would you like to provide a snapshot of what kind of distortion you are dealing with?

Naj

On Thu, Jul 23, 2015 at 7:00 AM, Barbara Kreilkamp <[log in to unmask]> wrote:
Dear FSL experts,

Could you please tell me if I'd  be able to use any retrospective distortion correction tools on my computed CBF maps? If yes, could you please point me to them?
My aim is to co-register these images to my structurals.

Thank you very much,
Best wishes,
Barbara



--
==================================
Najmeh Khalili-Mahani, PhD
Research Associate
McGill Centre for Integrative Neuroscience
Montreal Neurological Institute
McGill University

NW-143, 3801 University St.
Montreal, QC, H3A 2B4
Office: 514-398-1799






--
==================================
Najmeh Khalili-Mahani, PhD
Research Associate
McGill Centre for Integrative Neuroscience
Montreal Neurological Institute
McGill University

NW-143, 3801 University St.
Montreal, QC, H3A 2B4
Office: 514-398-1799