[VoxBo] Correlating t-values of different contrasts

[Daniel Kimberg]

David January wrote:
> I recently read a paper that argues that one needs to analyze
> patterns
> of overlapping activations via correlations from one contrast to the
> other.  Is there a way to do this in VoxBo?  If not, is there a way
> to
> convert the 3D cub files into 3D matrices that MATLAB can read?  The
> analysis strategy is to calculate t-values in an ROI an then
> correlate
> them with t-values from another contrast.

VoxBo doesn't have a facility for this.  Your best bet if you want to read cube files in SPM would be to convert them to analyze or nifti first.  The latter can be read into MATLAB structures with SPM5.  The former can be read by older versions of SPM, but the img file is just the 3D data.

dan