Whether or not you need to move the origin to the AC and reorient your
data will depend on many factors. Often it is not necessary, but
sometimes it can prevent disastrous misregistration.
Usually it is clear when the initial misregistration is poor. Either
things work well, or they fail completely. In your case, all seems to
have worked well. Changing the initial positioning of the images
should not change your findings.
Best regards,
-John
On 19 October 2011 17:45, Patric Hernandez <[log in to unmask]> wrote:
> To whom it may concern:
>
> While discussing segmentation steps with a colleague, a question was brought
> forward about making sure the rotation of the brain matches the MNI
> template, using the anterior commissure as the origin of the image. We came
> to the understanding that doing so minimizes the penalty term in SPM while
> maximizing mutual information.
>
> With that being understood, I wondered about my data, for which I used the
> DARTEL toolbox in my VBM processing. I did not do this step in my image
> processing. However, I have not had any issues with segmentation and have
> been very thorough about checking the quality of my images through each step
> in the process.
>
> I understand that aligning to the AC of the MNI template may stand to
> improve my data, but I am reluctant to reprocess all of my images due to the
> amount of time that will take to reprocess and reanalyze all the data for
> this large data set (n=88). My images and subsequent results have so far
> been useful and noteworthy. My understanding is that doing this additional
> alignment will not likely alter my results substantially, and if anything,
> should only improve the positive findings. So I find less reason to
> reprocess and would like to know the opinions of others with expertise in
> this area.
>
> Thank you,
> Patric Hernandez
>
> --
> MPA
> Neuroscience Doctoral Candidate
> UW-Madison
> [log in to unmask]
> [log in to unmask]
> www.patrichernandez.com
> Personal: (608) 772-9685
>
>
|