More than 50 researchers from local universities, health organizations and industry hit the gym at CAMH -- and simultaneously connected to global networks -- earlier this month.
They were taking part in CAMH’s first-ever BrainHack Toronto event. The Toronto site connected to a global, synchronized BrainHack conference spanning 36 cities and 16 countries.
Python...Github…brain imaging analysis – these topics might seem intimidating to some. But BrainHack delivered the latest news on tech and research innovations to researchers of all skill levels – from those learning fundamentals to more experienced hackers including post-doc students and scientists.
Dr. Erin Dickie, of CAMH’s Scientific Computing Working Group, and Baycrest’s Dr. Stephen Strother co-chaired this event, which was generously funded by the Ontario Brain Institute. The focus: Advances and Challenges in Neuroscience Data Integration.
“BrainHack is modelled on the tech sector hackathon but the hacking is cooperative rather than competitive,” notes Erin. BrainHack also emphasizes “intense, time-limited collaboration using open data and freely available software.”
Taking part in BrainHack Toronto were institutions including Baycrest, the University of Toronto, Holland Bloorview Kids Rehab, the Hospital for Sick Children, University Health Network, Sunnybrook Research Institute and CAMH, along with McMaster and Ryerson Universities.
On the agenda were topics including neural subtypes of depression and their implications for treatment, a look at open data sets, and the integration of cross-species imaging data. The Ignite speaker line-up included CAMH’s Dr. Leon French, and Dr. Jason Lerch of Sick Kids.
In addition to hearing from expert speakers both locally and via live-stream, participants took part in hacking sessions to work on interesting problems and case studies.
A number of participants also pitched their projects for the open hacking sessions on topics including:
- A dynamic brain-responsive Pomodoro timer for maximum productivity
- A library for connectome-based white matter atlas analyses
- A toolbox for automatic evaluation of predictive power of imaging biomarkers
- An R package for multivariate analysis resampling inference in neuroimaging
- The characterization of genes associated with an MRI-obtained cerebellar volume profile
- An algorithm for automatic segmentation of brain tumours from MR images
- Network alignment and brain variability.
This conference also featured the “Unconference,” says Erin. “We encouraged our participants to deliver what we called ‘Unconference Talks.’ These include demos of open-source tools or novel analysis methods they’ve written or become familiar with during their neuroimaging work.” The topics ranged from how to find, run and modify software to how to interpret results.
“BrainHack was a wonderful opportunity to exchange ideas and expertise with other neuroscience researchers,” says Dr. Yuliya Nikolova, Banting postdoctoral fellow at CAMH’s Campbell Family Mental Health Research Institute. “It allowed me to go outside my comfort zone and consider novel methods and perspectives to help move my research forward and create new collaborative projects.” Dr. John Griffiths, a post-doctoral fellow at Baycrest, agrees: “The relationships I made at BrainHack will be meaningful to my work – we applied upgrades to my project’s code that I would not have made, working alone.”
“This was a great opportunity to bring together researchers from multiple disciplines globally, to work together on innovative projects related to neuroscience,” says David Rotenberg, Manager of Scientific Computing in CAMH’s Information Management Group (IMG).