spm12 old seg failure

Submitted by zswgzx on
Hi all,
 
When I used spm12(v6225) spm_coreg + spm oldseg + spm oldnorm to normalize functional image to template space, several subjects' result didn't look reasonable. Here's how it looks like for one of them. Any help would be appreciated.

mean of fmri
https://drive.google.com/file/d/0B1-hPZ22Vq1VYWdleEc4MjN6U1k/view?usp=sharing

t1
https://drive.google.com/file/d/0B1-hPZ22Vq1VR2k3dzhGUUtSQTg/view?usp=sharing

The coreg result looks fine to me, but the tissue probability maps look weird:

coreg .mat file output
https://drive.google.com/file/d/0B1-hPZ22Vq1VZEFNSlVzNnROS1U/view?usp=sharing

GM tpm
https://drive.google.com/file/d/0B1-hPZ22Vq1VczFKa1hNOUstejg/view?usp=sharing
 
I tried to cut away some part of the neck and rerun, but it still didn't fix it. Any other suggestions?
 

Thanks!

Shengwei

 

YAN Chao-Gan

Fri, 09/18/2015 - 03:27

Hi,

One problem could be that the position of the T1 image is too far away from the standard template.

In DPARSFA, there are interactively reorienting steps, which could improve the starting position.

You can use DPARSFA to perform "spm_coreg + spm oldseg + spm oldnorm", simply check "Normalize by T1 unified segmentation", DPARSFA calls these procedures in SPM12.

Best,

Chao-Gan