Are there sources of test data (real or simulation) that are useful for comparing and contrasting different approaches to resting state (or rather, task independent) analysis procedures?

If so, what data?

If not, would it be a good idea, or even possible to have some such data available for common use?

In the previous forum thread, Daniel Marcus mentioned http://brainscape.org
It is a repository of resting state runs from multiple published studies.

It would be useful to know which if any of the datasets housed by the fmridc include resting data. Maybe if the readers of this forum divide it up, it wouldn't be too painful to go through all the studies. Fair et al. (2007, the ref is in the wiki for this group) also point out that you can do pretty well using the resting blocks from blocked BOLD studies.

The two data sets currently on Brainscape that would be most appropriate are from Fox et.al. 2007 Neuron and Fox et.al. 2005 PNAS. Both include control subjects, around a dozen each.

These data have all been run through the standard seed-region based correlation analysis engine on Brainscape. We'd be very enthusiastic to see others run these data sets through other tools.

We're also enthusiastic to incorporate additional analysis tools into Brainscape, if anyone is interested in seeing their tools distributed as server side pipelines.

-Dan

It may be of interest to this 'test data' discussion that the NYU_CSC_TestRetest Project (http://www.nitrc.org/projects/nyu_trt/) has just posted a news item regarding the release of a Test-Retest Resting-State fMRI Dataset (n = 25). See http://www.nitrc.org/forum/forum.php?for... for further information.