Brain Imaging in the Open
A spotlight on ME-BIDS, a 2018 Global Sprint project
Elizabeth DuPre (@emdupre_) is a doctoral student in the Montreal Neurological Institute at McGill University where she studies how brain networks interact in cognition, and what variability in the dynamics of these interactions can tell us about differences between people. Elizabeth was selected to join the current round of Mozilla Open Leaders with her work on ME-BIDS.
I interviewed Elizabeth to learn more about ME-BIDS and how you can help at the Mozilla’s Global Sprint 2018.
What is ME-BIDS?
ME-BIDS is a python package designed to improve processing of multi-echo functional magnetic resonance imaging (ME-fMRI) data. If we think of fMRI as taking pictures of the brain like frames in a movie, ME-fMRI takes multiple snapshots for each frame instead of only one. In processing this data, then, we can use these multiple snapshots to create high-quality, HDR images. Enhancing our level of detail is important, because it means that subsequent analyses can look across the ME-fMRI time series and use this extra information to decide which portions of the scene reflect brain activity, which is what we’re interested in!
Although there is an available processing pipeline for ME-fMRI data, it does not support the Brain Imaging Data Structure (BIDS) specification — a community-developed standard to ease sharing and re-analysis of fMRI data. ME-BIDS aims to create a BIDS App for multi-echo fMRI which will standardize and facilitate analysis of multi-echo fMRI data.
Why did you start ME-BIDS?
As a neuroscience graduate student, I have worked extensively with multi-echo fMRI data. This variant of fMRI was developed to reduce artifacts when acquiring fMRI data, such as those that arise when participants move their heads inside the MRI scanner. Because artifacts can impair our ability to interpret results from fMRI data, developing strategies to reduce their impact is important.
I believe it is equally important that the corrections that we apply to remove artifacts are transparent and well-documented, so that it’s clear to all users how these steps impact any conclusions drawn from the data. Initiatives such as BIDS are instrumental in achieving these goals, and, indeed, for non-ME-fMRI there already exists a BIDS App called FMRIPrep that ascribes to this “glass-box” philosophy. With ME-BIDS, I hope to create a similar tool for the pre-processing steps specific to ME-fMRI.
Why are you building ME-BIDS openly?
Although developing a new BIDS App for multi-echo processing will be valuable in ensuring the long-term viability of ME-fMRI, I believe the most important outcome of ME-BIDS will be the creation of a community for scientists and research teams using ME-fMRI. Science is fundamentally a collaborative enterprise, and the success of a method depends on the community supporting it. By developing openly, ME-BIDS will be better able to reach and receive input from existing ME-fMRI users, as well as to bring in new users and continue growing as a community.
What challenges have you faced working on this project?
We’re developing off an existing codebase, which gave us a great headstart but also emphasized the importance of things like readable code, licensing, and creating documentation. I hope that ME-BIDS will allow other researchers to join in efforts to improve and extend ME-fMRI processing directly, rather than forcing them to first invest in the time-intensive process of understanding undocumented code.
What kind of skills do I need to help you?
Everyone is welcome to contribute! Experience in programming with python would be particularly useful, though we are also looking for help in many other aspects of the project including documentation and logo design.
How can others join your project at #mozsprint 2018?
Check out our repository on GitHub and join us on Gitter Chat! During the sprint, we’ll be particularly focused on issues labelled as #mozsprint, but if you have other ideas or suggestions please open an issue! You can also contact me on twitter @emdupre_.
What meme or gif best represents your project?
Since Cohort C was almost #CatCrew…
Join us wherever you are May 10–11 at Mozilla’s Global Sprint to work on many amazing open projects! Join a diverse network of scientists, educators, artists, engineers and others in person and online to hack and build projects for a health Internet. Register today
