Dear Christian,
Thank you very much, I have updated to spm12 6685 and CAT12 r833, have
aligned the T2 to the T1 yet again using CAT12 and the 'y' warpfield
output looks more like what I should expect.
Thanks for your help,
Best wishes,
Barbara
On 17/01/2016 22:30, Christian Gaser wrote:
> Dear Barbara,
>
> your images were slightly misaligned, thus it is necessary to first co-register the T2 onto the T1 image (estimation is sufficient). However, you also have used a rather old version of CAT12 (VBM12) where the deformations were not correctly applied and some intensity issues happened. Please update to the newest release r833 and co-register your T2 onto the T1 image before applying the deformations.
>
> Best,
>
> Christian
>
> On Fri, 15 Jan 2016 16:54:49 +0000, Barbara Kreilkamp <[log in to unmask]> wrote:
>
>> Dear Marko and SPMers,
>>
>> Thank you for your fast reply, Marko, I have checked that the T1 and T2
>> are in exactly the same space, they align beautifully.
>> I have used the forward warp ('y') since I want to have the T2 in MNI
>> space native->Template; only when this did not provide the right result,
>> I tried the inverse - which was just a silly approach to see what would
>> happen :D.
>>
>> Funnily enough, the same data works just fine when using the forward
>> warp 'y' in VBM8/SPM8 which makes me wonder...
>>
>> After normalization I have not changed the T1-w image at all, the only
>> thing I notice is that normalize outputs a brain-extracted wmT1-w.
>>
>> I've uploaded the data which may make it easier to understand and I do
>> appreciate any help alot.
>> https://app.box.com/s/6m2vpiuqdko13kno3pvuqh6dil5skolx
>>
>> Thank you very much,
>> Best wishes,
>> Barbara
>>
>>
>>
>> On 15/01/2016 16:36, Marko Wilke wrote:
>>> Hi Barbara,
>>>
>>> a few things you may want to check: one, there is a push and a pull
>>> procedure to apply a deformation field, requiring the forward or the
>>> inverse deformation field. Getting this wrong may lead to strange
>>> results. Two, have you made sure that the T1 and the T2 are really in
>>> register with each other? If they are not lining up nicely in native
>>> space, then normalizing them with the parameters of one of them will
>>> not work. A related issue is if you moved the T1 to line up with the
>>> T2 after you computed the spatial normalization parameters, in which
>>> case the procedure you described will also not work.
>>>
>>> Good luck :)
>>> Marko
>>>
>>> Barbara Kreilkamp wrote:
>>>> Dear SPM-ers,
>>>>
>>>> I am wondering whether you could please help me sort this out, I've got
>>>> a non-brain-extracted T1w_acpc and T2w_acpc image in the same native
>>>> space after running T1w_acpc Segmentation (with MNI normalization) I
>>>> have saved both the forward and inverse deformation field, however, when
>>>> I apply either of these to the T2w_acpc image, the outputted wT2w_acpc
>>>> does not overlap with my wmT1w_acpc.
>>>>
>>>> Thank you very much,
>>>> Best wishes,
>>>> Barbara
>>>>
>>>>
>
|