[pymvpa] Classify neural across two very similar fMRI studies?

[Brian Murphy]

Hi,

indeed! Another study from our lab suggesting that it is worth trying:

http://repository.cmu.edu/cgi/viewcontent.cgi?article=1414&context=psychology
And there has been plenty of work on cross-participant prediction,
including across languages (training on speakers of one language,
testing on speakers of another),

Brian


On Thu, 2013-03-14 at 15:28 -0700, Jonas Kaplan wrote:
> Hi Danilo, 
> 
> We have done some work looking at classifiers trained on one sensory modality and tested on another modality (using pymvpa) to find sensory-independent representations, for example:
> 
> http://www.jneurosci.org/content/32/47/16629.abstract
> 
> In that paper the classification was done within individuals, but we have done it successfully across individuals as well (leave one subject out cross-validation).   There are many reasons why an analysis such as the one you propose might fail, but if learning transfers from one study to the other I think that is very interesting.  I would be interested to hear other opinions on this.  
> 
> -Jonas
> 
> ----
> Jonas Kaplan, Ph.D.
> Research Assistant Professor
> Brain & Creativity Institute
> University of Southern California
> 
> On Mar 14, 2013, at 11:42 AM, Danilo Bzdok <danilobzdok at gmail.com> wrote:
> 
> > Dear list,
> > 
> > I would like to apply pattern classification to a single seed region, in which two fMRI studies revealed convergence. Each of these fMRI studies used the same stimulus material across four tasks (to focus on top-down effects by subtraction). Both these fMRI studies used the same four tasks, but differed in the sensory modality on which the task is based (visual vs auditory). In every other respect the fMRI settings/protocoll were practically identical.
> > Here's now my question: Is it possible to train a classifier for neural activity patterns in the converging seed region to test whether activity in that area is sensory-independent or sensory-specific? Or is there too much confound when comparing between two subjects from very similar yet different fMRI studies? Any help would be greatly appreciated.
> > 
> > -- 
> > Best,
> > Danilo Bzdok
> > _______________________________________________
> > Pkg-ExpPsy-PyMVPA mailing list
> > Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> > http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
> 
> 
> _______________________________________________
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa


-- 
Brian Murphy
Staff Scientist
Machine Learning Department
Carnegie Mellon University
http://www.cs.cmu.edu/~bmurphy/