<div dir="ltr">Thanks a lot for the prompt reply!<div><br><div>I absolutely agree - clearly the variability between subjects makes a lot of sense. I select my ROI based on functional criteria, which means that (presumably) I take the voxels which process relevant information. I never get the overlap of these ROIs across subjects in search-light. I did not bring this point originally in order to simplify my question - I meant the ideal case where I have 10 identical twins in my group-level, with a complete ROI overlap :) But, even in this case I have a p-value issue which does not permit me to achieve significance in search-light. </div>
<div><br></div><div>BTW, why one should use a search-light which is substantially smaller than ROI? I deliberately make it of size of my ROI to make results more comparable between two analyses. I move my search-light each time one voxel in one of the directions, so each voxel participates in dozens of lights. At the end, I average all the predictions in each voxel. </div>
<div><br></div><div>I would appreciate if you notify me once your paper is in press.<br><br><div class="gmail_quote">On Mon, Feb 25, 2013 at 11:41 AM, J.A. Etzel <span dir="ltr"><<a href="mailto:jetzel@artsci.wustl.edu" target="_blank">jetzel@artsci.wustl.edu</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">On 2/25/2013 1:10 PM, Vadim Axel wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Absolutely naive question: suppose I have single a-priori defined ROI<br>
where I get a modest group-level beyond chance prediction of<br>
p-value=0.01 (one-tail t-test vs. 0.5, across subjects). Now I run a<br>
group level whole-brain search-light and I am expected to find at least<br>
one cluster of beyond chance prediction in the environment of my ROI.<br>
Correct?<br>
</blockquote></div>
No.<br>
<br>
I have a paper in (hopefully the last cycle of) review that goes into detail about these issues. But here's a brief version of some of the relevant ideas. I'm assuming you're using a linear SVM and proper cross-validation, and also that the searchlight is substantially smaller than the ROI.<br>
<br>
Two possible explanations come to mind:<br>
1) The single searchlights are too small to hold enough voxels to classify accurately, but the ROI can, because there is weak information present in much of the ROI. Linear SVMs can combine weak information from many voxels, so can sometimes classify better with more voxels.<br>
<br>
2) There is a lot of spatial variability between subjects. Suppose only a small part of the ROI is informative. If that part falls withing the ROI for everyone, then the ROI might classify well at the group level. But if each person only has a small informative area on their searchlight map, the group map could come out non-significant (people's maps don't overlap enough).<br>
<br>
<br>
A few suggestions:<br>
1) If your hypothesis is about the ROI, stick with the ROI-based analysis, adding control ROIs (or whatever) as necessary, but not doing the searchlight analysis.<br>
<br>
2) If you need the searchlight analysis for a particular purpose, do some sensitivity testing, and look closely at the single-subject maps. For example, how much do the maps change with different searchlight radii? Did you normalize to atlas space before or after the searchlight? Did you smooth the data? Smooth the individual subject maps? etc.<br>
<br>
3) Check the sensitivity of the ROI-based finding. For example, How much does it change if the ROI boundaries are altered slightly? How much variation is there between subjects - does the ROI classify well in most everyone, or just a few people?<br>
<br>
<br>
Hope this gets you started, and good luck.<br>
Jo<span class="HOEnZb"><font color="#888888"><br>
<br>
<br>
<br>
-- <br>
Joset A. Etzel, Ph.D.<br>
Research Analyst<br>
Cognitive Control & Psychopathology Lab<br>
Washington University in St. Louis<br>
<a href="http://mvpa.blogspot.com/" target="_blank">http://mvpa.blogspot.com/</a><br>
<br>
______________________________<u></u>_________________<br>
Pkg-ExpPsy-PyMVPA mailing list<br>
<a href="mailto:Pkg-ExpPsy-PyMVPA@lists.alioth.debian.org" target="_blank">Pkg-ExpPsy-PyMVPA@lists.<u></u>alioth.debian.org</a><br>
<a href="http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa" target="_blank">http://lists.alioth.debian.<u></u>org/cgi-bin/mailman/listinfo/<u></u>pkg-exppsy-pymvpa</a><br>
</font></span></blockquote></div><br></div></div></div>