BrainHack projects

Phys2bids is Physiopy's BIDSificator, covering a good span of physiological vendors data formats already. One major missing feature, however, is the lack of support for some MRI vendors physiological data format.
On top of that, the tutorial section of the documentation could be improved to simplify usability, and user testing could help us detect new features.
Within this project, we propose different aims for all types of skills, learning goals, and collaborations:

1. Implement data support for SIEMENS [coding skills required: python]

2. Implement data support for hd5f data [coding skills required: python]

3. Implement a feature to automatically find groups of TRs in physiological files [coding skills required: python]

4. Update the documentation to help better usage of phys2bids [no coding skills required]

5. Update the "tutorial" section of phys2bids with more practical examples and videos [no coding skills required, a light knowledge of restructured text is welcome]

6. User test phys2bids to find feature that could be added [requires running phys2bids on data and opening issues on phys2bids' github pages]

Useful links: https://github.com/physiopy/phys2bids

Background: fMRI and NEM (EEG, MEG, ECoG/iEEG) form two complementary modes of assessing the brain. Common knowledge suggests that fMRI has better spatial resolution, while EEG and MEG have better temporal resolution. Both modalities are needed to understand the brain!

Recent years have seen an extensive expansion of open science efforts, particularly in the fields of psychological and cognitive neuroscience. In particular, open science datasets that can be downloaded and used by researchers worldwide have become one of the primary avenues for conducting open science, maintaining a high level of transparency, and addressing the reproducibility crisis.

Currently, most of the efforts in the open science neuroimaging community are directed towards the fMRI modality. However, as stated above, NEM is an inseparable part of open science.

Project aim: To develop a tool that tags NEM and fMRI open datasets in various repositories together. Using this tool, members of the open science community will obtain suggestions for NEM datasets using similar paradigms when they are interested in using an fMRI dataset (and vice versa). Thus, researchers using open science to study a specific aspect of brain activity will immediately and conveniently have access to both these inseparable aspects of the brain under this condition – the time sensitivity and the spatial sensitivity. The researcher would be able to analyze these two datasets together and gain a deeper understanding of the brain.

Background: Currently, tools, concepts, and procedures of open science, especially in the fields of psychology and neuroscience, are not commonly applied. As was clear from the current workshop, there is a clear audience and appetite among young Taiwanese students to engage more with open science practices.

However, there are numerous obstacles to implementing open science practices in the Taiwanese higher education system. One of the main obstacles to such an implementation is the pushback on open science practices from administrators, institutional ethical boards, university management, funding agencies, and governmental bodies.

One way to address the discrepancy between grassroots-level interest in open science and higher-level objections to these practices is to actively lobby (in the political sense of the word) among these bodies for adopting an open science-positive approach.

Project aim: The student lobbying company will seek ways to actively engage with high-level institutional and governmental officers to promote an open-science-friendly approach (meetings, letter campaigns, advertising, etc.)

Want to learn more about git, set up your local machine (e.g. laptop) for computational work or create your own free website using GitHub?

Great, this is the Brainhack tutorial for you! Just find @PeerHerholz any time during the Brainhack and bring your computer.

Nilearn is a fantastic `python package` for all things MRI. It also includes a lot of tutorials and examples. However, some of them are maybe missing a bit of information, especially for folks who are just starting to use this `package` or doing MRI data analyzes. Thus, it might be nice to go through some sections and evaluate if they should be updated.