Hi Steve
Thanks for this information. I found (as you might expect) selection of
the target FA map is critical for good registration. The mean FA maps
and skeleton now look fine. Just for comparison I would like to try and
run flirt in tbss_3 instead of the affine transformation, but I am not
sure what lines of code to change (if any) in the following step.
echo "affine-registering target to MNI152 space"
${D}/areg $M target.hdr -dofout target_to_MNI152.dof -parameter
${D}/affineCC.par
${D}/transformation target.hdr target_to_MNI152.hdr -dofin
target_to_MNI152.dof -linear -target $M
Any help would be much appreciated.
Cheers
Stephen
On 27/09/2006, at 6:24 PM, Steve Smith wrote:
> Hi - I'm not sure what you mean by "unusual distortions" - I'm
> guessing that you're just referring to the fact that the target image
> is being upsampled to 1mm for use as the "standard space", but it it's
> primary intrinsic directions aren't aligned with the image matrix
> axes. You might be able to solve this by using flirt to align the
> target image to mni152 (with 1mm output) and use that as the target
> instead - hopefully that should work ok (flirt is more robust than the
> affine transformation used in tbss_3).
>
> Cheers, Steve.
>
>
>
> On 27 Sep 2006, at 08:24, Stephen Rose wrote:
>
>> Hi
>>
>> I am trying to analyse some preterm neonate DTI data using TBSS. I
>> seem to be running into
>> problems with tbss_3_postreg, with registration of my neonate brain
>> FA maps to mni152 space.
>> My thoughts are that it would more appropriate to complete the entire
>> analysis in raw image space
>> (eg target FA map) rather than forcing registrations into 152 space.
>>
>> In a previous query by someone else it was suggested to change:
>>
>> $FSLDIR/bin/flirt -in $FSLDIR/etc/standard/avg152T1_brain -ref
>> $FSLDIR/etc/standard/avg152T1
>> -out MNI152 -applyisoxfm 1
>>
>> to
>>
>> $FSLDIR/bin/flirt -in target -ref target -out MNI152 -applyisoxfm 1
>>
>> I have tried this option but get unusual distortions on the
>> target_to_MNI152 and FA skeleton maps
>> (see attached files).
>>
>> Is there a way to bypass the registration step to 152 space (in
>> tbss-3-postreg) and keep
>> everything in the target image space for the remainder of the
>> analysis steps? This seems the only
>> option until someone comes up with a neonate equivalent of the
>> mni152. Any help would be much
>> appreciated.
>>
>> <skeleton.jpg>
>> <target.jpg>
>
>
> -----------------------------------------------------------------------
> ----
> 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
> -----------------------------------------------------------------------
> ----
>
>
Dr. Stephen Rose
Senior Research Fellow
Centre for Magnetic Resonance
University of Queensland
Brisbane, QLD, 4072 Australia
Telephone: +61 7 32327480
Mobile: 0412023958
Fax: +61 7 33653833
|