POOR COREGISTER

Submitted by Ilka on Fri, 01/03/2014 - 20:56

Hi,

which settings underly the coregister function in DPRSF?

We did the coregistering with SPM using the default settings. To date when coregistering our fMRT data we always got decent results. This time when coregistering our resting state data with the same settings we usually used in  7.5%  of the cases the results were really poor. Any ideas what could be the problem?  Should we use different coregister settings than in the normal fMRT data?

Many thanks

Ilka

Submitted by ZHANG_RESTadmin on Wed, 01/08/2014 - 12:24

I'm not familiar with your experiment background and with what's your resting-state fMRI scanning parameter resembles your task fMRI one.  But usually two fMRI datasets use the same coregistering method and setting.  If the "coregistering" you mentioned is "normalization" in SPM, then I think you'd better further check your normalization settings.  Because there are lots of normalization ways, I don't know what the exact problem is as long as I don't know how you normalized your data.   Usually, if your funtional MRI data did not roughly align with the EPI template prior to normalization, the normalization algorithm will fail to reach good solution, producing bad normalization result.   So you may need to reorient your resting-state fMRI data to roughly match the position and orientation of the EPI template by using SPM: display.