[FieldTrip] cluster statistics on dynamic ROI

[Haris Tzagarakis]

Dear Fieldtripers,
I have a question about cluster-based statistics at the source level 
using an ROI. The option cfg.roi exists in ft_sourceanalysis but 
requires the use of atlas anatomic labels.
My case is slightly different in that I have an ROI defined dymamically 
by a prior source level analysis and I want to evaluate source level 
statistics only within the
voxels of that ROI for a different parameter.

I ended up creating volumes with the appropriate dimensions and setting 
all the volume array values for the parameter I want to test to NaN 
except for the voxels of the ROI. I also set up the "inside" field to 
contain the index values of the ROI voxels only.
This seems to work when using cfg.correctm    = 'cluster' (which is what 
I want) but was wondering if anyone can see a potential problem that I 
might be missing?

With Best Wishes
Haris

-- 
Charidimos [Haris] Tzagarakis MD, PhD, MRCPsych

University of Minnesota Dept of Neuroscience

office:
Brain Sciences Center
Minneapolis VA Medical Center
Tel:612-467-1363