search

WhitakerLab / BrainsForPublication

Code and documentation for the Brains for Publication proposal for the 2016 OHBM Hackathon
MIT License
31 stars 23 forks source link

Online rendering of statisical maps in NeuroVault #6

Open KirstieJane opened 8 years ago

KirstieJane commented 8 years ago

@chrisfilo and the NeuroVault development team are also exploring visualisation tools through the NeuroVault platform.

You can easily create 3D (pial, inflated or flattened) figures using pycortex when you upload your maps to NeuroVault as well as interactive volume visualisation using papaya.

NeuroValut volume visualisation

An example of a volume visualisation from http://neurovault.org/images/16208

NeuroValut volume visualisation

Corresponding surface visualisation from http://neurovault.org/images/16208/pycortex

You can also generate glass brain images for each map, such as this one:

NeuroVault glass brain
KirstieJane commented 8 years ago

@chrisfilo: I absolutely adore these visualisations, but I'd like to have a discussion (with maybe some expert and non-expert NeuroVault-ers) around the barriers that people may meet regarding using them.

To give an example, if I were in the middle of doing analyses (rather than finished and sharing the final maps) I'm not sure I'd want to bother uploading all my maps just to make a figure to share with my research group at our next informal meeting.

What do you think? Are there ways of streamlining this process?

chrisgorgo commented 8 years ago

The advantage of putting "work in progress" maps on NeuroVault is that you can easily share them with collaborators who are not necessarily in the same room (think of emailing a link). Additionally, maps are stored on NeuroVault servers so it easier to recover them if something happens to your computer. On top of that we have formal reverse inference tools such as neurosynth decoding and image search which can help you interpret your findings in the context of existing literature. Collections can also be private if you don't want others to (yet) see your maps. Summing up I think there are some great advantages of using NeuroVault even for analyses that are still ongoing.

Nonetheless, you are right that if all you want is a static image uploading an image could be seen as a burden. It would not be that hard to provide papaya as a client side viewer with no need for uploading files (but maintaining the advantage of no need for installing software), but it's not really the long term mission of NeuroVault. We want people to share their maps and giving them tools to visualise and decode them is an incentive to start using NeuroVault early in the lifespan of the study so when the manuscript will be submitted/published maps will already be on NeuroVault (because you submitted them earlier to share them with collaborators, visualise them or decode them).

I hope this helps!

KirstieJane commented 8 years ago

I'm convinced! I think the only visualisation that I'd like to see for offline use is the glass brain one. But I think your point about sharing the link with collaborators etc is a really good one.

Is there documentation in NeuroVault that promotes this use case of uploading early but not making public for a while? If not then I think we could make some as part of this project :smile:

chrisgorgo commented 8 years ago

You are right - we should describe this "workflow" more explicitly in the documentation!

On Sat, Apr 16, 2016 at 8:56 AM, Kirstie Whitaker notifications@github.com wrote:

I'm convinced! I think the only visualisation that I'd like to see for offline use is the glass brain one. But I think your point about sharing the link with collaborators etc is a really good one.

Is there documentation in NeuroVault that promotes this use case of uploading early but not making public for a while? If not then I think we could make some as part of this project [image: :smile:]

— You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub https://github.com/KirstieJane/BrainsForPublication/issues/6#issuecomment-210845866

KirstieJane commented 8 years ago

Good stuff - ok - lets make that one of the goals for this project (rather than changing any code from NeuroVault for offline use) and if the documentation looks good then it can be merged into NeuroVault too :smiley:

chrisgorgo commented 8 years ago

Sounds like a plan! I will not be at the hackathon, but I can contribute this ahead of time.

On Sat, Apr 16, 2016 at 9:22 AM, Kirstie Whitaker notifications@github.com wrote:

Good stuff - ok - lets make that one of the goals for this project (rather than changing any code from NeuroVault for offline use) and if the documentation looks good then it can be merged into NeuroVault too [image: :smiley:]

— You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub https://github.com/KirstieJane/BrainsForPublication/issues/6#issuecomment-210849303

KirstieJane commented 8 years ago

Brilliant - I've made an issue asking for feedback on where the best place for documentation is (#3) so feel free to push is in whatever direction best suits you as a remote contributor!

Any bits of instructions that you can contribute would be great, but don't worry about it too much - I'm sure there will be someone willing to put the directions together from all the current NeuroVault documentation.