Hello Fatemeh,
Thank you very much for your upload - we have looked carefully at the undistorted images and the directionality is the same in both cases. The "streaks" in the shift map are only due to a change in how we display the shiftmap ( we now display the extrapolated map outside the brain rather than the masked version ).
Hope this Helps
Kind Regards
Matthew
> Many thanks Matthew for looking into this problem. Since my first email to
> you, I
> have managed to resolve some of my problems with filed map and BBR: The
> main issue was that FSL 5 needed the non-BETed magnitude image for
> Unwrapping as well. However I still have 2 concerns as stated below:
>
> The UPLOAD REFERENCE is 715510. The file named R1.tar.gz is the FLIRT
> registration and field map
> correction with FSL4. The file BBR.tar.gz is the same design.fsf file ,
> run with FSL5 using BBR and field map correction. The 2 outstanding issues
> for me are now:
>
> 1) why does the 'unwrap shift map,in voxels' ('EF_UD_shift+mag.png'), have
> streaking outside the brain when unwrapping in FSL 5 compared to FSL4? I
> have used the same input field maps and magnitude images.
> 2) The unwrapping direction seams to be different in FSL 5 ( to my
> untrained eye), even though i use +y for both.
>
>
> Many thanks and best wishes
> Fatemeh
>
>
>
> Dr Fatemeh Geranmayeh
> Wellcome Trust Clinical Research Fellow
> The Computational, Cognitive and Clinical Neuroimaging Laboratory
> Imperial College London
> Telephone : 020 7594 7994
>
>
>
>
>
> On 08/11/2012 15:11, "Matthew Webster" <[log in to unmask]> wrote:
>
>> Hello Fatemeh,
>> Can you tar up your 2 feat directories and upload them
>> to:
>>
>> http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi
>>
>> and let me know the 6-digit reference number?
>>
>> Kind Regards
>>
>> Matthew
>>> Dear experts
>>>
>>> I loaded a design.fsf file from a previously run analysis in FSL4 that
>>> contained field map correction and FLIRT registration into FSL 5. I then
>>> re ran the same first level analysis with FSL 5 and BBR option (not
>>> making
>>> any other changes, for example no changes to B0 unwrapping
>>> option),however
>>> it appears that the field map correction makes the distortions much
>>> worse
>>> when used in FSL 5. (see attached files). I have also attached the
>>> design.fsf file.
>>>
>>> I checked the analysis with a different patient who had less atrophy,
>>> but
>>> the problem persists.
>>>
>>> Any advise will be much appreciated and I am happy to send more files
>>> (there was a limit of 200 K on this email)
>>>
>>> Many thanks
>>> Fatemeh
>>>
>>>
>>>
>>> Dr Fatemeh Geranmayeh
>>> Wellcome Trust Clinical Research Fellow
>>> The Computational, Cognitive and Clinical Neuroimaging Laboratory
>>> Imperial College London
>>> Telephone : 020 7594 7994
>>>
>>>
>>>
>>>>
>>>
>>> <field map FSL 4.pdf.zip><design.fsf_FSL5.zip><field map _FSL 5
>>> .pdf.zip>
>
|