Display of significant cluster

[Miguel Ángel Rivas Fernández]

Dear all, 

I performed a GLM in SPM and I don't obtained significant results at peak level using the FWE correction. However, two my results survive at cluster level and I obtained two significant clusters with the FWE correction (p<0.05).

How can I display only these two significant clusters using the xjview? 

Best regards, 

[Xu Cui]

You may "select" the two clusters and then "pick" in xjview.

Xu

----------------------------------------------------------------
Xu Cui, Ph.D
Department of Psychiatry and Behavioral Sciences
Stanford University
Blog: http://www.alivelearn.net

Profile: http://www.linkedin.com/in/cuixu
----------------------------------------------------------------

--
You received this message because you are subscribed to the Google Groups "xjView discuss group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to xjview-discuss-g...@googlegroups.com.
To post to this group, send email to xjview-dis...@googlegroups.com.
Visit this group at https://groups.google.com/group/xjview-discuss-group.
For more options, visit https://groups.google.com/d/optout.

[Miguel Ángel Rivas Fernández]

Dear Xu,

Thank you very much for your help. I have an additional question.

I selected one significant cluster and then pick in the "Pick/Cluster info" button. Now the program show to me information about this cluster such as the brain regions that are included in this cluster as well as the number of voxels located in the gray and white matter.

My question is, what brain regions should I report in my study? I am not sure but I suspect that I should to report those that have the parenthesis (aal). However, I have noticed that there is no correspondence between the number of coordinates that appears in the SPM result table (16) and the number of "aal" regions that xjview shows for this cluster (28).

This could be due to the fact that the SPM results table only shows 16 local maxima more than 4.0 mm apart?. In that case, what criteria does xjview take into account when shows this information?

I send to you information that show to me xjview program as well as a screenshot of the SPM results table.

Cluster 1
Number of voxels: 1378
Peak MNI coordinate: -15  35   8
Peak MNI coordinate region:  // Left Cerebrum // Frontal Lobe // Sub-Gyral // White Matter // undefined // undefined
Peak intensity: 3.8388
# voxels    structure
 1378    --TOTAL # VOXELS--
  984    White Matter
  829    Right Cerebrum
  723    Frontal Lobe
  524    Left Cerebrum
  476    Sub-lobar
  303    Sub-Gyral
  256    Gray Matter
  248    Extra-Nuclear
  171    Inferior Frontal Gyrus
  149    Limbic Lobe
  115    Anterior Cingulate
  114    Medial Frontal Gyrus
  114    Caudate_R (aal)
  101    Cerebro-Spinal Fluid
  101    Lateral Ventricle
   97    Frontal_Inf_Tri_R (aal)
   96    Caudate
   96    Corpus Callosum
   84    Middle Frontal Gyrus
   78    Caudate_L (aal)
   71    brodmann area 10
   70    Frontal_Med_Orb_R (aal)
   52    Frontal_Inf_Oper_R (aal)
   52    Caudate Body
   51    Insula_L (aal)
   50    Frontal_Sup_Medial_R (aal)
   44    Superior Frontal Gyrus
   44    Caudate Head
   43    Frontal_Inf_Tri_L (aal)
   37    Cingulum_Ant_L (aal)
   35    Frontal_Mid_R (aal)
   31    Precentral_R (aal)
   30    Cingulum_Ant_R (aal)
   30    Frontal_Med_Orb_L (aal)
   30    Cingulate Gyrus
   25    Inter-Hemispheric
   24    brodmann area 32
   23    Insula
   13    brodmann area 9
   12    Frontal_Sup_Medial_L (aal)
   10    Frontal_Mid_Orb_R (aal)
    9    brodmann area 24
    9    Frontal_Sup_R (aal)
    8    Precentral Gyrus
    8    Frontal_Sup_Orb_R (aal)
    8    Insula_R (aal)
    7    brodmann area 13
    7    brodmann area 46
    6    brodmann area 45
    5    Parietal Lobe
    5    Pallidum_L (aal)
    4    Putamen_L (aal)
    3    Inferior Parietal Lobule
    3    Cingulum_Mid_R (aal)
    2    Putamen_R (aal)
    2    brodmann area 25
    2    brodmann area 6
    2    Frontal_Sup_Orb_L (aal)
    2    brodmann area 33
    1    Subcallosal Gyrus
    1    Frontal_Inf_Orb_R (aal)
    1    Olfactory_L (aal)
    1    Olfactory_R (aal)
    1    brodmann area 47
    1    Lentiform Nucleus
    1    Frontal_Sup_L (aal)
    1    Claustrum

Sorry for this simple question and thanks in advance.

Best regards,

Miguel

--

Miguel Ángel Rivas Fernández

[Xu Cui]

Miguel,

In the report, the number in the 1st column is the number of voxels in the region labeled in the 2nd column. The number of rows has nothing to do with the number of local maxima in SPM reports.

In terms of choosing the anatomical names, it's totally up to you. Usually you should pick one which is specific enough but also at the same time contains a significant number of voxels. AAL labeling is in general a good choice.

Xu 

[Miguel Ángel Rivas Fernández]

Dear Xu,

Thanks for your help. As you said in the prevoius message, the xjview report gives the number of voxels of brain regions labeled in the second column (including the aal regions). However I was wondering if is there any way to exportación automatically the significant aal regions and its associated MNI coordinates?

Many thanks in avance

Best

[Xu Cui]

Unfortunately xjview can't do this.

Xu

[Miguel Ángel Rivas Fernández]

Dear Xu, 

I used your useful script cuixifindstructure.m script and to localize the significant brain regions obtained in my SPM results. However, some of these regions are located in white matter or undefined locations:

// Right Cerebrum // Frontal Lobe // Superior Frontal Gyrus // White Matter // undefined // Frontal_Sup_Medial_R (aal)'

//Temporal Lobe // Superior Temporal Gyrus // undefined // undefined // Temporal_Pole_Sup_R (aal)'

Should I consider these results as "not significant"? otherwise, should I only report in my study those regions located in gray matter regions?.

Apologies for this simple question. 

Best regards, 

[Xu Cui]

It's common that some of the voxels of a significant cluster are located in white matter. You may check the stats of the voxels in each region (xjview can do this), and use an appropriate anatomical name for your cluster.

Xu

[Miguel Ángel Rivas Fernández]

Hi Xu, 

That is, for example, if I have a cluster in the superior temporal gyrus but the script cuixufindstructure say that is located in white matter:

// Left Cerebrum // Temporal Lobe // Superior Temporal Gyrus // White Matter // undefined // Temporal_Sup_L (aal)

I can check in the nearest surrounding voxels if there is a significant voxel in this same cluster that is located in a gray matter region. In that case, I would have to report the MNI coordinates and its associated t statistic that give to me the xjview (see the attached screenshot). It is ok?

 // Left Cerebrum // Temporal Lobe // Superior Temporal Gyrus // Gray Matter // brodmann area 22 // Temporal_Sup_L (aal)

Thank you very much for your help but I would like to be sure that I have understood your recommendations.

Best, 

[Xu Cui]

Usually, you should pick the peak voxel (the one with the highest T value) in a cluster.

Also, in your case, I think you can try to lower your p-value so that your cluster size is small enough, and then click "pick cluster/info" button.

btw, the bands in your image (top left glass view) looks suspicious. I don't know the cause but you may want to look into it.

Xu

----------------------------------------------------------------
Xu Cui, Ph.D
Department of Psychiatry and Behavioral Sciences
Stanford University
Blog: http://www.alivelearn.net

Profile: http://www.linkedin.com/in/cuixu
----------------------------------------------------------------