Hi,
1. There is no need to use fslsplit to split the files into 30 separate
volumes, just use fslmaths with the three 30 volume 4D files (fslmaths
<file1> -add <file2> -add <file3> -div 3 <averagefile>).
2. Merging all three averages into a big 4D file and running eddy_correct
should correct for any between average motion that occurred. If there
wasn't any, then you shouldn't see any difference between those two methods.
Peace,
Matt.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Markus Gschwind
Sent: Wednesday, October 31, 2007 11:24 AM
To: [log in to unmask]
Subject: Re: [FSL] Averaging diffusion image
Thanks Matt, great!
1. Following your suggestion I produced that big4Dfile, I resplit
(fslroi) it into 3 files and now I'd like to split them each into the
30 volumes in order to average them.
When trying to split it I get an error message (unlike I did when
splitting analyze-files):
[fsl@localhost Esplit]$ fslsplit -t data1.nii.gz
** ERROR (nifti_image_read): failed to find header file for '-t'
** ERROR: nifti_image_open(-t): bad header info
Error: failed to open file -t
Cannot open volume -t for reading!
Why?
2. I did both ways:
A. I run eddycurent on each 4D(session) and then fslmerged them to
a big4Dfile
B. I first fslmerge them to a big4Dfile and then run eddycurrent
over the whole.
At eye control in fslview there is no difference between the both outputs.
I substracted the outpouts one from another (fslmaths -sub) and the
differecnce was empty!
Is that what you expect?
Cheers and thanks!
Markus
Quoting Matt Glasser <[log in to unmask]>:
> Hi,
>
> 1. You should only have to run eddy_correct once.
>
> 2. I have run into this problem also, which is why you should verify the
> output of eddy_correct (particularly for newer versions, older versions
> didn't seem to make as many mistakes). In some cases, I have forgone
using
> eddy_correct on a dataset if the output is worse than the input.
>
> 3. use:
> fslroi <big4d> <1staverage> 0 30
> fslroi <big4d> <1staverage> 30 30
> fslroi <big4d> <1staverage> 60 30
>
> Peace,
>
> Matt.
> -----Original Message-----
> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
> Of Markus Gschwind
> Sent: Wednesday, October 31, 2007 7:20 AM
> To: [log in to unmask]
> Subject: Re: [FSL] Averaging diffusion image
>
> Hi Matt (again! ;-))
> Thanks for your clarifications.
>
> 1. So you run edyycurrent several times over the same data. Is this
> not a problem?
>
> 2. After I fslmerged the three 4D files to one and run eddycurent, I
> still find that in the movie function of fslview there are still
> motions visible (the same or even worse as in the non eddycorrected
> file)
>
> 3. Could you please give me a hint how to use fslroi to re-separate
> them (there are 3x30 volumes in this big volume).
>
> Thanks a lot!
> Cheers,
> Markus
>
>
>
> Quoting Matt Glasser <[log in to unmask]>:
>
>> Hi,
>>
>> I usually run eddy_correct on the whole dataset before averaging by
>> concatenating the 3 averages into a single big 4D file, with fslmerge,
and
>> running eddy_correct on that. Then split the 3 averages back out using
>> fslroi and you can use fslmaths as you suggest below.
>>
>> Peace,
>>
>> Matt.
>>
>> -----Original Message-----
>> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
Behalf
>> Of Markus Gschwind
>> Sent: Tuesday, October 30, 2007 6:26 AM
>> To: [log in to unmask]
>> Subject: [FSL] Averaging diffusion image
>>
>> Hi Experts!
>>
>> I have 3 aquisitions of diffusion data D1, D2, D3 for one subject.
>> I would like to take the average image for further processing.
>> For that purpose I used the fslmaths -add and -div commands.
>>
>> My question:
>> Actually I would have to correct for eddycurrent BEFORE averaging, is
>> that true?
>>
>> Thanks a lot for help!
>> Markus
>>
>>
>> --
>>
>> Markus Gschwind, M.D.
>> Laboratory for Neurology and Imaging of Cognition
>> Dept of Neurosciences
>> University Medical Center (CMU)
>> 1 Michel-Servet - 1211 GENEVA - CH
>>
>> Tel 0041 (0) 22 379 5324
>> Fax 0041 (0) 22 379 5402
>> email: [log in to unmask]
>> http://labnic.unige.ch
>>
>
>
>
> --
>
> Markus Gschwind, M.D.
> Laboratory for Neurology and Imaging of Cognition
> Dept of Neurosciences
> University Medical Center (CMU)
> 1 Michel-Servet - 1211 GENEVA - CH
>
> Tel 0041 (0) 22 379 5324
> Fax 0041 (0) 22 379 5402
> email: [log in to unmask]
> http://labnic.unige.ch
>
--
Markus Gschwind, M.D.
Laboratory for Neurology and Imaging of Cognition
Dept of Neurosciences
University Medical Center (CMU)
1 Michel-Servet - 1211 GENEVA - CH
Tel 0041 (0) 22 379 5324
Fax 0041 (0) 22 379 5402
email: [log in to unmask]
http://labnic.unige.ch
|