Segmentation failure for T1 coregistered to tilted fMRI

Hello some structural scans failed segmentation and normalization after coregistering to RS fMRI scans that had been tilted forward (30 degrees on the x axis to avoid inferior-frontal dropout).  We used the unified segmentation method.  It seems SPM may not be able to match the structural to the standardized tissue templates if the orientation differs a lot.  We were able to resolve the problem somewhat by reorienting the images.  Has anyone else experienced this? 

We attempted newseg and DARTEL but the resulting images were striated (striped) along the x axis. 

DPARSFA has proven very user-friendly for batch processing.  Thank you. 

-Cody

Forums:

 Have you tried using DAPRSFA to do the coregistration and segmentation? You may need to check reorient options both for rs-fMRI data and for structure image before coregistration. It will work.

Han Zhang

 Hi thanks for your reply.  I did use DPARSFA for coregistration and segmentation.  Coregistration worked but T1 segmentation failed.  The RS fMRI images are tilted 30 degrees on the x axis.  This seems to have caused problems in segmentation of the coregistered T1 image.  Segmentation provides the normalization matrix used for the RS fMRI.  Has anyone else experienced this problem?  How could I work around this problem using DPARSFA?  Thanks again, Cody

Hi,

Why not use "reorent Fun*" to correct the tilt in your functional images?

Please do "reorient Fun*" and "reorient T1*" seperately.

Best,

Chao-Gan