Hi,
we are using the gre_field_mapping sequence.
I found a very helpful script online:
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/epidewarp.fsl
The results I get with this look reasonable. Still I have to find what I
did wrong...
Something very strange I encountered with preparing my fieldmap
previously: the phase ranges from 0 to 4095, so to rescale to -pi to pi,
I would think that I should divide by 2048, subtract 1, multiply by pi.
However, when doing this, the phase ranges from -3pi to pi. Simply
dividing by 4096 and multiplying by pi gives the correct range from -pi
to pi... Is there an explanation for this? I only have one dataset right
now, so I don't know if it is just this dataset or a general problem.
Cheers,
Kerstin
@Andreas: Ich bin im Lehrstuhl Haase/Jakob gross geworden :) Wir sind
euch hier uebrigens 8 Stunden voraus
Andreas Bartsch wrote:
> Hi Kerstin,
>
> what kind of sequence are you using - a double gradient vs. spin echo fieldmapping? In any case, you should just use the echo time difference, not twice the value.
> Schöne Grüße zurück, es muß spät bei Dir sein-
> Andreas
> PS: Hast Du in Würzburg studiert?
>
>
> -----Ursprüngliche Nachricht-----
> Von: FSL - FMRIB's Software Library im Auftrag von Kerstin Pannek
> Gesendet: Mo 22.09.2008 07:51
> An: [log in to unmask]
> Betreff: Re: [FSL] AW: [FSL] AW: [FSL] fugue distortion correction on dti stretches CC
>
> Hi Andreas,
> you are right, I managed to make another mistake, that "fixed" the
> problem - my bad, it worked only once...
> Still I think the trouble is when I generate the fieldmap. Is it correct
> that asym is 2x echo time difference? I read this somewhere, and it
> seems to work, but maybe just another mistake that cancels out another -
> I seem to be good at that today...
> Gruesse ins schoene Frankenland!
> Kerstin
>
> Andreas Bartsch wrote:
>
>> Hi Kerstin,
>>
>> are you sure? I would definitely NOT expect a large difference to be introduced by an echo time difference error of 0.04ms.
>> Cheers-
>> Andreas
>>
>> ________________________________
>>
>> Von: FSL - FMRIB's Software Library im Auftrag von Kerstin Pannek
>> Gesendet: Mo 22.09.2008 02:59
>> An: [log in to unmask]
>> Betreff: Re: [FSL] AW: [FSL] fugue distortion correction on dti stretches CC - solved
>>
>>
>>
>> Hi all
>> It turned out the problem was in the generation of the fieldmap itself.
>> The echo times I used were inaccurate - OsiriX (not the current version)
>> reported echo times of 7.40 and 4.90, while ImageJ reported 7.38 and 4.92.
>> Cheers
>> Kerstin
>>
>> Kerstin Pannek wrote:
>>
>>
>>> Hi Andreas
>>> Using --unwarpdir=y- is actually what I did for this subject,
>>> resulting in the overstretched CC. The size and shape of the brain
>>> seem correct when doing so. When I don't use --unwarpdir=y-, the whole
>>> image is distorted even more.
>>> Cheers
>>> Kerstin
>>>
>>> Andreas Bartsch wrote:
>>>
>>>
>>>> Hmmh, maybe your have negative blip encoding? A least you could try
>>>> to reverse the sign for y.
>>>> Cheers-
>>>> Andreas
>>>>
>>>> ________________________________
>>>>
>>>> Von: FSL - FMRIB's Software Library im Auftrag von Kerstin Pannek
>>>> Gesendet: Fr 19.09.2008 09:02
>>>> An: [log in to unmask]
>>>> Betreff: [FSL] fugue distortion correction on dti stretches CC
>>>>
>>>>
>>>>
>>>> Hi all!
>>>> I would like to use fugue to correct distortions on my dti data using a
>>>> field map.
>>>>
>>>> I did the following:
>>>> 1) eddy_correct dti data
>>>> 2) extract and bet b0 image
>>>> 3) bet magnitude, mask fieldmap with betted magnitude
>>>> 4) warp magnitude (fugue -w)
>>>> 5) flirt -dof 6 warped magnitude to betted b0
>>>> 6) apply flirt transform to masked fieldmap
>>>> 7) unwarp dti data (fugue -u)
>>>> 8) dtifit
>>>> 9) flirt -dof 6 FA to structural (T1)
>>>>
>>>> I found that the corpus callosum in the FA map (and corrected b0 image)
>>>> is "stretched" when compared to the structural image (see attachment).
>>>> The screenshot shows the structural image, overlaid in red with FA
>>>> (thresholded at 0.3).
>>>> Does anyone know what might cause this problem, and how I could fix it?
>>>> Thanks for your help!
>>>>
>>>> Cheers
>>>> Kerstin
>>>>
>>>>
>>>>
|