This is a really cool website that automatically synthesises fMRI activation maps from a couple thousand studies onto a brain image for you. Taking a couple of clicks and less than a minute, it's quick, easy and could be very useful. This post will give you a quick overview of using this tool.
Starting from the homepage (found at http://neurosynth.org/) click the term-based maps from within the images section.
...and voila! A map of voxel activation's found from within their database that are associated with the entered search terms.
Alternatively, you have the option to enter image co-ordinates and the inverse will happen, so the probabilities of those co-ordinates being associated with various areas and functions will come up. Using [x,y,z] +44 +4 +8 will show you a high probability of insula activation, .78 for example.
I have only just found this tool but I think the utility of it is clear. A really quick way to look up a region that you may find active in your fMRI data. It is this sharing of 'public access' tools that is really commendable for those people sharing their cool software inventions in the imaging field, and with the promise for the ability to compose 'full-blown' meta-analyses in the upcoming months, they will probably help others.
There is one thing that I do need to ask someone though. Concerning the nature of the blobs being made by the site. Does the intensity of the blobs match up to the # of voxels reported as active, or do they reflect the magnitude of the BOLD effect? Or perhaps reflect both aspects? This is something that I am not too sure about, and think would need to be clarified before guiding researchers when finding unexpected results. And what about papers that don't explicitly report there functional co-ordinates, can they also be incorporated into this sites database?