[Marsbar-users] Re: Otto: MarsBaR problems

[Matthew B. <mat...@uc...>]

Hi,

> Using MarsBaR I am getting some strange errors for when picking some 
> of the regions, although many regions are working. The error message 
> is below. Any idea why this might happen? Shall I download the region 
> file again?
> 
> *****************
> Design reporting                        :                        ...done
> 
> Mapping files                           :                        ...done
> Calculating globals                     :                        ...done
> Fetching data                           :                         1/10
> Warning: No valid data for roi 1 (Amygdala_L)

This means that every voxel in the Amygdala is missing in at least one
of the functional scans you selected.  This can be the field of view of
the acquisition, or problems with misregistration of one or more images
to the others, or to the template.  Check your mask.img in the SPM
results for this design to see if it includes the amygdala, for example.
> Also, how can I extract summary information (such as mean and SD) for 
> regions? I tried Data->Export Data, but I can get only the time 
> course for the regions. 

If you select the option to save the summary data, the time course is
the time course of the mean voxel value in the region for each scan. 
There is a thread in the marsbar mailing list containing some code to
get the variance across voxels in the ROI, but this is a little bit more
complicated.

> Running a second level analysis, I would like 
> to extract the average Z-score in the contrast image for each 
> individual subject.

Well, you can get the average t statistic per subject, just by selecting
the t images, one for each subject for the comparison you are interested
in, in the extract data (full) option, and then saving the summary
(mean) data out to file etc.

Best,

Matthew