fMRI quality control

Summary:

With the emergence of large fMRI databases, there is a need for tools for high-throughput quality control. Although a few tools (such as mit-art) exist to detect outliers in fMRI databases, most standard packages propose no solution besides the visual inspection of individual results. This project is to build a work group with the following objectives:
1. Document existing strategies implemented in a variety of labs.
2. Compare the outcomes of these strategies on public databases such as the 1000 functional connectomes.
3. Brainstorm for novel approaches.
What can I do ? 
There are at this stage two ways to contribute. First, if you are developing/using a package which includes some tools to screen large or small fMRI databases for quality control, you can share a few slides that summarize your strategy. Second, if you used such an approach on a large public database such as ADHD200 or the 1000 functional connectome, please prepare a summary of the outcome, such as the one that was generated for the NIAK release of the preprocessed ADHD200 database.
How can I join ?

Please contact pierre.bellec [at no spam] criugm.qc.ca.
Will this project be represented at Brainhack 2012?

During Brainhack, we will meet to prepare a summary of the similarities/differences between the various approaches and test if the different strategies gave consistent results. If enough material is gathered, the results will be the basis of a panel discussion to define a consensus strategy and debate about the limitations and possible developments in this area.