Just out of curiosity: Are there any plans to extend Neurosynth also to sMRI?
Ahoi hoi @JohannesWiesner,
as far as I know/can tell neurosynth does that already.
In the honest and nice words of @tal on the FAQs page: “The parser is dumb. It considers anything that looks like a coordinate in a table to be a relevant activation. It doesn’t know what the activation means.”. Therefore, the parser also does not distinguish between functional and structural data, as it looks for coordinates in tables. So if the publication includes that and the search term you’re interested in, it should work.
For example, if you check the results for the accumbens search term and within that search for structural you’ll see some sMRI studies that were included.
Or was your question more related to sMRI as modality specific Neurosynth adventures?
Pinging @tal here for further input and to intervene in case I’m off.
HTH, cheers, Peer
A structural version of Neurosynth is feasible, and it’s come up many times over the years. It would certainly be nice to have, and the structural literature is probably big enough to support it now; it’s mostly a matter of activation energy. At the moment the whole Neurosynth platform is undergoing a massive rewrite (that probably won’t be done for at least another 6 months), but if someone wants to adapt the (eventual) code to the structural setting, I’d certainly support that.
(Note that while it’s true that there are indeed some sMRI studies in the Neurosynth DB, heuristic filters are used to screen out structural studies, so there at least shouldn’t be a lot of them.)
1 Like
Thanks for the clarification @tal, sorry for missing the heuristic filters! My insufficient neurosynth knowledge strikes once again…
That being said, I would be up for helping re the structural setting as a sincere apology.
Sorry, I don’t check NeuroStars very often—only just saw this! If you’re still interested, I’d be happy to chat (probably after the course is over). No worries if not, seeing as it’s been 7 months. 
No biggie! I guess it would be a useful feature to introduce something like a modality filter/argument/etc.? Given the recent developments this should happen over at nimare, eh?