Analysis of common regions by XJVIEW

[lion gao]

Dear Xjview expert,

 

I run a common regions dispaly  in Xjview, and found a nice result.. While i get a bit confused when i try to describe the procedure-- is there reference i can quote for common regions analysis? and  usually, what's the underlying reason for common regions analysis.  It seems common region is not the same as conjunction analysis, nor it is simply another combine d group analysis. BTW, I use (p < 0.01 ) when doing common regions analysis.

 

Another problem i meet is that when i get the statistic table in Matlab, I find that the T Value from the volume button of xjview can be over 100! in common regions analysis, but usually it should be 3 ~8.  so what's the meaning of this T value?

 

These problems are quite important for my thesis, I would appreciate a lot if you can help in any way :) Thanks very much in advance!

 

Best wishes,

Gao

[Xu Cui]

Hi, Gao,

First, good luck with your thesis.

The "common region" function is very simple: at a given threshold (say
p=0.001), xjview will find the supra-threshold voxels for each image
and then find a subset of these voxels which are common to all images.
I personally use this function to compare if two contrasts give out
similar regions, and if yes, what are the overlapping regions.

I don't know if there is a formal reference you can cite about common
region as it is so simple and straightforward. As I never did
conjunction analysis, I don't know how to compare them.

In xjview, the values in the common region are the product of values
from individual images. I use this to roughly identify which voxel is
more likely to be "common" but there is no solid reason behind.

Hope this helps,
Xu

--
----------------------------------------------------------------
Xu Cui
Department of Psychiatry and Behavioral Sciences
Stanford University
Blog: http://www.alivelearn.net
Try Peaya Paper, a paper/pdf manager http://www.peaya.com
----------------------------------------------------------------

[lion gao]

Dear Xu cui,

 

Thanks for the answer, I see that it's simply common regions among contrasts above some threshold.

Another issue a bit strange is that why the T-value generated (by pressing volume...etc) is so high ( can be 20 ~ 100), but actually they are range from 4~7 when check individually.

 

Best Wishes,

Gao

2009/11/18 Xu Cui <cuixi...@gmail.com>

[Xu Cui]

the values in the common region are the product of values from individual images. I use this to roughly identify which voxel is more likely to be "common" but there is no solid reason behind. I will simply change the values to 1 in the future version to avoid confusion.

Xu

[lion gao]

Ya...  not familiar with these explanations.

Anyhow,  Thanks for your prompt Answer :)

2009/11/19 Xu Cui <cuixi...@gmail.com>