Bootstrapping seed-based functional connectivity

[TheChymera]

SAMRI can now generate seed-based connectivity plots, however, the seed regions of interest might be at slightly different positions and of slightly different shapes/sizes between subjects, at least.

A useful feature may be to bootstrap the seed region, so that that it can iteratively grow into the region making the most functional sense at that approximate location in each animal.

In the following plots, for instance, the first subject shows a purported dorsal raphe functional region filling up the entire seed region and extending somewhat ventrorostrally, while the second subject shows an incomplete filling of the seed region at the dorsocaudal end, and a slight ventreal extension. Better covering these regions in each animal would likely allow us to extract better signal for the connectivity analysis.

@damaggu any ideas on how to best accomplish this? after the seed-based connectivity analysis is performed, the image is available as an array-like object. Program-logic-wise, this should be really easy to do, in that we only have to run an algorithm on that array to get n improved seed region, which we can then pipe right back into the functional connectivity function.