XJVIEW INTRODUCTION: an alternative to viewing results in SPM
 
1) Type xjview at the matlab command prompt: >>xjview
 
2) File – select Open Images
 
3) Go to “File – Open Images”
 
4) Navigate to QA demo, Top Down, Sample One directory .
 
5) Select the spmT_0006, (as opposed to SPM.mat). This is a 2-back WM > vigilance task *Note both task positive and task negative activations.
 
6) Superimpose on high resolution single subject brain.
 
7) Change threshold to .01: Note that you can easily move thresholds without reselecting Tmap!!
 
8) Navigate to Rt parietal (BA40).
 
9) “Pick Cluster” – show # voxels in parietal cluster and locations of those voxels.
 
10) Superimpose on Mask.img
 
Note: Although the data originally looked good – that is you see the typical WM network. However, when you superimpose this on the mask, you see that the back of brain is missing.
-
-
 
11) Load Tmap from Sample 2 WM and repeat steps above.
 
12) Run Display_rois either via the BIT GUI or by typing the following on the matlab command prompt: >> display_rois.
 
13) Select the SPM canonical single subject high resolution brain: SPM8/canonical/single_subj_T1.nii.
 
14) Number of Images to Display? Select “TWO”.
 
15) Navigate first to the Topdown Sample 1 directory and select the mask.img and then navigate the the Topdown Sample 2 directory and select that mask.img.
 
16) Color Image? Select “RED” for the first mask and “BLUE” for the second mask.
-
-
 
Note: The purple are represents the common voxels of these two masks. Hence these two subjects alone eliminate any possible activation outside this purple region in any 2nd level group analysis.
 
Note: You could also use SPM’s Imcalc to take intersection of the two masks. To do this you would enter the two mask images and enter the following expression i1 .* i2. Also, you can use a program called GlmMask to bypass SPM’s masking routine and instead implement a user specified mask. However you may be ignoring artifacts if you choose to do this.
 
Susan Whitfield-Gabrieli MAR09