Dear Dana,I'm not sure what Christine is talking about with respect to problems or confusionsbetween FSL and AFNI. Both packages were heavily involved in the developmentof NIFTI and interpret things equivalently. I suspect that your problems are notrelated to any incompatibility between FSL and AFNI and I am not aware of anythat exist (there were some for the first few months of NIFTI support many yearsago, but it has not been the case for some time now).I'm confused about the fact that your EPI to Structural is bad but your EPI to MNIis good. Are you using the FSL GUIs? The EPI to MNI alignment is not independentbut a combination of the EPI to Structural and the Structural to MNI. So you shouldonly get a good EPI to MNI result if the other two are also good. If the EPI to Structuralis poor then this should also show up in the EPI to MNI. If you are not using ourGUIs then please let us know a bit more about the steps you are running.If you have any over 90 degree differences in initial alignment then things can gettricky and you are probably better off running fslreorient2std on all of your imagesfirst. This will get them all in roughly the same orientation as the MNI. The 30 degreetilt of your EPI acquisition should not be a problem and we would normally not doanything special to try and fix it.I'm not familiar with the AFNI programs being talked about here, but if they areonly modifiying the qform and/or sform information then I don't see that this wouldbe problematic for FSL.If you are still having problems then please feel free to upload some examplesto our upload site at:I think that the mask problem is likely to be a symptom of other problems and so,as Steve also suggested, I think it is the fundamental problems that we need toaddress.Hope this helps.All the best,MarkOn 5 Oct 2011, at 14:58, dana eldreth wrote:Hi Steve and Christine,Thanks so much for your responses. As Steve suggested, I rechecked the BET images and BET did a really good job on this pilot subject. The registration of EPI to MNI and Structural to MNI are good, but the EPI to Structurals are not good. It doesn’t register when I use the original oblique files, but it works when I do the 3dwarp in AFNI before any of the processing steps in fsl. However, when I put things into the higher level analysis the sum of all masks is wrong. I tried removing the header information prior to processing the data in fsl as Christine suggested, but that didn’t appear to make a difference either in the higher level analysis (same problems with sum of all masks). Is there anything else that I could try?
Thanks so much!!!DanaOn Wed, Oct 5, 2011 at 9:41 AM, Christine Zakrzewski <[log in to unmask]> wrote:
Hey. I would try removing oblique information from the header before reading your warped images with fslview. This can be accomplished with AFNI program 3drefit; from the command line-3drefit -deoblique <input> . (Also bear in mind the negative determinant of the template matrix in fsl can create some confusion—especially with NifTI files being swapped between fsl and afni.)
Good luck!
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Stephen Smith
Sent: Wednesday, October 05, 2011 4:56 AM
To: [log in to unmask]
Subject: Re: [FSL] oblique data registration
Hi - it's very unusual for FLIRT not to be able to get a good registration between EPI and structurals - so I would concentrate on resolving why this isn't working robustly. Were the structurals correctly brain-extracted when fed into FEAT registration?
Cheers.
On 4 Oct 2011, at 14:57, Dana wrote:
Dear FSL Experts,
I’ve acquired pilot subject data (Anatomical and Functional) in the oblique plane at 30 degrees above AC-PC to minimize signal dropout in frontal regions due to susceptibility differences. I realigned the data to the main structural highres image and standard MNI space, but it was way off even if I did full search. I looked at the data in fslview and the labels were off because the brain is tilted 30 degrees. Therefore, I decided to warp the data into the cardinal plane using AFNI’s 3dWarp and Tshift programs (functional) and the data labels were accurate in fslview. I have multiple runs, so I put each run into a separate first level analysis and then combined the feat files into a 2nd level analysis. When I looked at the sum of all input masks, I knew that something went wrong because the mask covers the whole square for each brain slice rather than just the brain. As far as I can tell the registration looks really good for each run. I’ve seen this problem before, but that was due to fsl not being able to understand the labels provided in the conversion program I was using so I know that’s not the problem here since the labels are accurate. Just to see what would happen, I ran the oblique data through the whole processing stream and the sum of all masks looks right, but the registration is completely wrong. It seems like the 3dWarp program may be doing something to the data that’s creating this error. If anyone has any suggestions of how best to analyze oblique data in FSL, I would greatly appreciate it. I’ve looked it up, but I’m not finding too much.
Also, this may be a related problem, but this is the first time I’ve analyze multiple runs in FSL and I notice that the brain activation for main effects (task vs baseline fixation) doesn’t always produce a lot of activation (not even in the visual cortex and this task has IAPS pictures). I’m wondering if this could be related to the problem I’m having here or if it may be a separate issue. Any insight into this issue would also be greatly appreciated.
Many thanks,
Dana
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
The information in this transmission is intended for official use of the Kessler Foundation. It is intended for the exclusive use of the persons or entities to which it is addressed. If you are not an intended recipient or the employee or agent responsible for delivering this transmission to an intended recipient, be aware that any disclosure, dissemination, distribution or copying of this communication, or the use of its contents, is strictly prohibited. If you received this transmission in error, please notify the sender by return e-mail and delete the material from any computer.