 |
 |
Jess, I believe that the non-zero voxel issue has to do with the fact that for the combined runs, the images have been put into a standard space. Therefore, you should be getting the same # of voxels per subject per area when you are looking at the non-thresholded z-stat image. Then again it might be solely due to the fact you are just analyzing the z-stat image. In other words, you are taking a mask from the atlas, which has a specific number of voxels, and overlaying that on a portion of the image which has non-zero voxels. So every time you do this you will get the same number of voxels after masking. You get values for mean % signal change on the individual runs because there are many timepoints (TRs) for each voxel; hence, it can calculate this. The results for the averaged data has done just that, it has averaged the results for each run leaving you with only 2 "time-points." Hope this helps, Carlos >I have been carrying out some featquery analyses on some data. I have >noticed that when I run featquery on data averaged over two runs of a task >(each participants gfeat) the number of non-zero voxels in the output report >seems to be exactly the same for all participants for the PE, COPE, Tstat and >Zstat. Very rarely there is a participant where the number of non-zero voxels >differs from all the rest of the participants. It only differs between the >participants for the thresholded z stat. However when I run featquery on >each of the individual runs of the task not only do I get values for mean % >signal change that do not average to give the means I get when I run the >analyses on the gfeats but the number of non-zero voxels differs quite a lot >between participants for the PE, COPE, Tstat and Zstat. > >I wondered if anyone could explain why this is and if I am doing something >wrong. > >Many many thanks > >Jess