Frequently Asked Questions

1. What is this website for?

The purpose of this website/database is to collect and distribute statistical maps of the brain. Such maps are acquired by scientists all around the world using brain imaging techniques such as MRI or PET in a combined effort to map the functions and structures of the brain. Unfortunately, due to the limitations of the academic paper form (as in publication not pulp) in which most findings are disseminated, only coordinates of peaks are reported. This makes doing meta-analyses and hypothesis-driven studies very difficult. With unthresholded statistical maps, one can find subthreshold effects that are consistent across studies, which cannot be achieved using just peak coordinates. You can read more about NeuroVault in our paper.

2. Why would I submit anything?

When you submit data to NeuroVault you will be provided with a permanent link for each map. The links lead to an interactive visualization. You can share you results with other researchers without sending files around! Readers of your papers will also appreciate additional clarity provided by an interactive visualization. Therefore you can include links to your NeuroVault dataset in your paper (see this and this paper). On top of that, all the maps uploaded to NeuroVault can be decoded using

3. How do I know my data is safe and will be available for the years to come?

All of NeuroVault is backed up daily into off site storage. In addition all of the public collections that have been associated with a paper (through the DOI field) will be archived in Stanford Digital Repository (SDR). This operation is performed in bulk twice a year and ensures long term preservation of maps deposited in NeuroVault.

4. How does decoding work?

Decoding is an experimental feature that takes your map and compares it with patterns associated with cognitive terms in thousands of papers. All the hard work is provided by, which will pull your map from NeuroVault and run spatial correlation restricted by a gray matter mask. Give it a try - just hit the "Decode" button above any image. This can potentially be very useful in interpreting your results. The feature is still experimental and we would be grateful for any feedback.

5. How can I cite NeuroVault?

If you refer to NeuroVault in an academic context (manuscript or talk) please cite our paper:

Gorgolewski KJ, Varoquaux G, Rivera G, Schwartz Y, Ghosh SS, Maumet C, Sochat VV, Nichols TE, Poldrack RA, Poline J-B, Yarkoni T and Margulies DS (2015) a web-based repository for collecting and sharing unthresholded statistical maps of the brain. Front. Neuroinform. 9:8. doi: 10.3389/fninf.2015.00008

Citing this paper is the best way to show your appreciation and will help us with securing funds necessary to run and improve NeuroVault.

6. Is it only about fMRI?

No - not at all. We accept any brain data that is stored in a 3D NIFTI format and has been transformed into MNI space. This means that structural studies (for example VBM) as well as PET are more than welcomed here.

7. How can I access the database in a programmatic way? Is there an API?

Yes there is a RESTful API - you can find it here. There are also python and R clients for the API. Please let us know if you build something cool using NeuroVault!

8. I found a bug. How can I report it?

Excellent! Please use the "Give feedback" button in the toolbar.

9. Who is behind the project?

NeuroVault is being built by a group of dedicated developers led by Chris Gorgolewski. They include (in no particular order): Tal Yarkoni, Yannick Schwartz, Camille Maumet, Vanessa Sochat, Joe Wexler, Daniel Margulies, Anton Burnashev, Joke Durnez and Asier Erramuzpe. Additionally many people have been consulting on this project, including Tom Nichols, Yaroslav Halchenko, Michael Hanke, Russell Poldrack, Michael Milham, Satrajit Ghosh and the INCF Neuroimaging Data Sharing group. The project is supported by International Neuroinformatics Coordination Facility and Max Planck Society (in particular Max Planck Institute for Human Cognitive and Brain Sciences and Gesellschaft für wissenschaftliche Datenverarbeitung mbH Göttingen).

10. I'm a developer and I have some ideas on how to improve the website.

Fantastic! You can send us a pull request on github.

11. Sharing statistical maps is great, but I would like to share whole datasets!

NeuroVault is dedicated to just statistical maps, but you should check out OpenNeuro. Additionally you might consider publishing a data paper to get credit for your sharing efforts.

12. Those are some beautiful viewers! Where did you get them from?

We use Papaya and pycortex for 3D view.

13. What license is the data distributed under?

All data in NeuroVault (with exclusion of private collections) is distributed under CC0 license.

14. My image looks perfectly aligned with the MNI template in FSLView but when I upload it to NeuroVault it's misaligned in the web viewer.

Use this command line FSL tool: fslcpgeom MNIatlas image_with_weird_header. (Many thanks to Leonie Lampe for the tip).

15. Should I add a NeuroVault link to the paper I'm about to submit? What's the best practice of doing this?

Linking unthresholded maps is a great way to allow the readers of your paper to explore your results and reuse them in their work. We recommend grouping all maps from one study/paper in a single collection and including a link to that collection in your manuscript. Additionally one can include links to individual maps in figure captions. An example can be found in the following publication, the caption of Fig. 3. Please note that if you decide to use a private collection (e.g., during peer-review), a unique string is added to the URL so people would not be able to guess the address. Additionally, it's worth mentioning that if your collection was initially private (so it got longer URLs) and you later made it public, both URLs (public and private) will resolve correctly. This allows users to keep collections private during peer-review and make them public after the paper gets published (which is not required but can be preferred by some).

16. The original data was in MNI152 space. How were the surface maps obtained?

The surface maps were obtained with the registration fusion approach. If you use the surface maps in a manuscript or talk, please reference the paper:

Wu J, Ngo GH, Greve DN, Li J, He T, Fischl B, Eickhoff SB, Yeo BTT (2018) Accurate nonlinear mapping between MNI volumetric and FreeSurfer surface coordinate systems. Human Brain Mapping. doi: 10.1002/hbm.24213

All transformations (including the reverse fsaverage-to-MNI transformations) are available here.