| I posted this question a few days ago, with no response. I'm trying my luck
again.
|
| When I display various files in the results directory using "Display" in
SPM99, the various files such as RPV, ResMS, beta_0001, con_0004, spmT_0005 and
mask ALL seemed to be masked images. All data outside the brain appears black.
|
| However, when I use "Results" to statistically interrogate results, the 3
orthagonal images aren't masked and there are numerous non-zero voxels outside
the brain.
|
| Why?
I don't know. Can you superimpose your results on the mask image using
orthogonal views to make sure that the results shown in the glass brain
correspond to those shown on the orthogonal views?
|
| Where/how is the "mask.img" that is created when specifying a model used?
The mask.img is made very crudely:
The "global" intensity within the brain is computed as being the
average intensity of all those voxels that are more than an
eighth of average intensity of the whole image.
A threshold based on this "global" is then used to determine which
voxels are analysed. To be included in an analysis, all voxels
in a time series should have intensities greater than this threshold.
The mask.img is just a representation of those voxels included in
the analysis.
Best regards,
-John
|