Hi Nick,
This is certainly possible, but not having touched ToyTalk in ages, I am not actually sure how reliable it still is.
At the moment, the ToyTalk implementation is mostly matlab code that duplicates some of what is done in the C-api that AFNI/SUMA use. I am wondering if we should make the effort to start using the same API from matlab and/or from Python. Otherwise it will be too much effort to keep up the various APIs. As it stands in the AFNI pacakge, we currently exchange surfaces, surface-based datasets, volumes, coordinates, tracts, etc. It would be nice to have the same functionality be available for MATLAB/Python/etc. We've done something comparable for R which now loads all things loadable by AFNI via the same set of C functions for opening datasets and that has made R I/O matters a lot easier.
Let's continue this discussion over email if you're interested in doing something at that level. It SHOULD not be too hard :) . Otherwise, I can add a few small changes to the current ToyTalk tools to get you the information you want. we can also discuss this at OHBM or ISMRM should you be attending this year.
cheers,
z
