-
Notifications
You must be signed in to change notification settings - Fork 0
CPSG Grant Application
Brainhack is an event designed to bring together enthusiasts with various levels of technical ability and interest in the brain to learn from each other and generate collaborative projects. The event will span several days consisting of technical workshops, project proposals, and copious amount of open hacking (where participants will group together and work on their projects). The projects can include anything from artistic representations of brain data to developing novel analysis methodologies to creating basic tools/practices helping with analysis. Creativity of the participants is the only limiting factor towards project generation. One project that will be a part of the event will increase the productivity of individuals collecting neuroimaging data. Currently, labs that collect neuroimaging data do not have a standard (or a well documented standard) for organization. Too much time is lost interpreting poorly organized data. A published standard (Brain Imaging Data Structure or BIDS) for organizing imaging data exists and it has been adopted by labs across America and around the world. There hasn’t been an initiative like BIDS before, making the only competition for this standard the idiosyncratic (and often unwritten) rules adopted by labs. Brainhack will provide an excellent venue to distribute this standard and create a pathway towards adoption at the University of Iowa, improving our current research practices and productivity.
Brainhacks evolved from the need for more flexibility than traditional conferences to give participants opportunities to work on collaborations and/or learn skills in an interactive setting. Starting in 2012, Cameron Craddock and his colleagues at the Open Neuro Bureau created Brainhack events to bring together brain enthusiasts from a variety of backgrounds to work together in open collaboration. The outcomes of these events include open software, data sharing initiatives, and novel solutions to data acquisition barriers (Craddock et al. 2016).
Since the first Brainhacks in 2012, the events have proliferated to different sites around the world gaining more participants and creating more collaborative endeavors. Last year, (2017), there were 36 Brainhack events spanning 16 countries with over 1000 participants (Brainhack Global). Currently, there are 37 sites signed up to host a Brainhack event again this year (including the University of Iowa), demonstrating the continuing popularity of these events.
Brainhacks combine elements from hackathons, unconferences, and educational workshops. Hackathons are project oriented open collaboration events, where participants with the requisite skill-sets contribute to projects they are interested in. Unconferences can include informal presentations of ideas, projects, or panel discussions. Educational workshops give participants necessary tools to perform open science and enhance their own analytical pipelines. Combining all three elements gives participants an opportunity to interact with the event and their peers regardless of their current skill level.
Labs and departments that collect data from the research scanners at the University of Iowa are in dire need of a conference like Brainhack. The current data management practices on how to handle and organize the massive quantities of data coming from the scanners are lab specific (or sometimes specific to particular members within the lab). In addition these organization standards are not well documented (if at all). This makes sharing data between labs or re-analyzing an old data-set from several years ago a nightmare. The number of productive hours lost to understanding how others organized their data and rewriting analysis tools to accommodate individual data organization schemes is innumerable. Furthermore, these unwritten individual organization schemes invite errors in analysis whether it's a mis-specification of a file or missing meta-data required for a particular analysis.
This is not a problem unique to the University of Iowa, and in response to this need, an open standard has been published and has been adopted in labs across America and around the world (Gorgolewski et al. 2016; Poldrack 2017). This data standard is well documented and has taken input from a large number of researchers, and is still being expanded and discussed today. The standard is named Brain Imaging Data Structure, or BIDS for short. Adopting BIDS at the University of Iowa will improve productivity of researchers and support the growing need for open science practices (e.g. exhaustive data reporting and sharing). There is already a software ecosystem supporting BIDS if researchers choose to use those tools, or they can continue using their own analysis tools.
(include location, participants, and community served)