Thanks John and Volkmar for your feedback.
The problem I'm having is with the distortions. I've tried MI and AIR's
alignlinear, both of which gave me suboptimal results. Is there a way to
use spm's normalize function to perform some nonlinear warping using the
mprage as the target and the b0 as the source? I'm not sure if it will
handle cross-modality registration.
Mike
John Ashburner wrote:
>>I'm acquiring a DTI sequence and a high resolution 3D MP-RAGE sequence,
>>and I would like to coregister the FA maps to the individual mprage
>>scans for some structural segmentation work.
>>
>>Is there a way to register the T2-weighted B0 images to a T1 weighted
>>MP-RAGE structural scan with reasonable accuracy?
>
>
> I would suggest trying the coregister button. Mutual Information (and other
> information theoretic methods) should do a reasonable job.
>
> One possible source of error would be distortions in your DTI data.
> Coregistration only uses a rigid-body model, so does not account for the
> distortions.
>
> Best regards,
> -John
>
--
Michael A. Yassa, B.A.
Neuro-imaging Research Technologist
Division of Psychiatric Neuro-imaging
Department of Psychological and Brain Sciences
Johns Hopkins School of Medicine
600 N. Wolfe St., Phipps 322
Baltimore, MD 21287
Tel: 410-955-7861 / Fax: 410-614-3676
E-mail: [log in to unmask]
Web: http://pni.med.jhu.edu/people/yassa/yassa.htm
|