Dear Kyle,
Concerning 1), the tutorials you looked at were probably based on older SPM versions, in which the "Mean" column was placed at the end of design matrix. You just have to adjust your contrast vectors accordingly so that they match your design matrix.
Concerning 2), the Brodmann map is not the best way to localize clusters, I'd suggest to turn to one of the probabilistic and/or multi-subject anatomy-based atlases/labels (n30r83, LPBA40, Neuromophometrics, Harvard-Oxford) or alternatively, to the Anatomy toolbox if you want to rely on cytoarchitectonic labels. Leaving this aside, to get an impression of the data you can overlay the SPMs on the tissue probability maps you selected for segmentation and/or the MNI templates, and I'd also create a mean image of your smoothed, segmented files (e.g. wc1 or mwc1). This way it should be easier to judge whether the clusters are reasonable or not. Note that in case of VBM analyses, one usually defines an absolute threshold to restrict the analysis to voxels with a certain amount of grey matter (during design specification, Masking \ Threshold masking would have to be set to Absolute, go with something like 0.1 or 0.2 to incorporate voxels whose GM intensity exceeds the threshold in all your subjects). Alternatively and in case of rather severe atrophy you might also go with the procedure discribed by Ridgway et al. (2009, http://dx.doi.org/10.1016/j.neuroimage.2008.08.045 ), also see http://www0.cs.ucl.ac.uk/staff/g.ridgway/masking/
Best
Helmut
|