Open software for neuroscience

Brainstorm advances fundamental and preclinical electrophysiology research

Source: Sylvain Baillet, The Neuro

A new open science software app developed by my colleagues and I will help neuroscientists analyze, visualize and report on ever-increasing volumes of data, for free.

Twenty years ago, a young post-doctoral fellow at the University of Southern California (USC) came up with the idea that for sophisticated methods in neuroscience to have a real impact in research labs, their trainees and the industry, they had to be wrapped up into well-documented, user-friendly software easily accessible to everyone.

Brainstorm was born out of this idea, initially to serve the vast and diverse community of researchers and students interested in applying imaging techniques to electro- and magnetoencephalography (EEG, MEG). Fast forward to today, and Brainstorm now has 24,400 registered user accounts worldwide and has produced approximately 1,200 research articles, many in leading publications such as Science, Neuron, and Nature Neuroscience.

In addition, my collaborators and I have put 300 pages of online documentation, gigabytes of tutorial data, and delivered 40 training workshops around the world, from Chili to Japan, to about 2,000 research professional and trainees. Some of our users in Iran have organized their own training sessions independently. We have an online forum with tens of thousands of messages from users supporting other users with how to best use the software, responding to questions and suggesting new features and improvements to the user experience: it’s truly open science in action, every day, for everyone.

Overall, the impact has surpassed our expectations. Every occasion to read from researchers using our software or to hear from trainees in the workshops we organize drive our motivation to move forward with the approach of free, open-source software development and sharing.

Brainstorm is now setting another important milestone via the publication in the open-access Nature group journal Scientific Data on Oct. 25 2019, of a significant extension of the app’s features to a whole new realm of neuroscience data.

After adding magnetic resonance imaging, intracranial EEG from patients, and near-infrared spectroscopy (this latter via a collaboration with Christophe Grova’s group at Concordia University), the software now includes a new portfolio of tools and entire sections of online documentation specifically to respond to the needs of the very large and diverse community of electrophysiology researchers. This field is one of the largest in neuroscience, and the present reality is that labs need to develop their analytic software tools themselves to advance their research.

In addition to Richard Leahy, my longstanding collaborator at USC, I have teamed with electrophysiology specialists Adrien Peyrache and Chris Pack at The Neuro to produce the new, extended app, with the core developments contributed by Konstantinos Nasiotis, an Integrated Program in Neuroscience (IPN) graduate, and Martin Cousineau and François Tadel, two Brainstorm core software developers in my team.

Research funding for this work was provided by the National Institutes of Health and an Innovative Ideas grant from the Healthy Brains for Healthy Lives program at McGill. We are looking now at creative solutions to ensure the long-term sustainability of the project, possibly via the creation of a software foundation, inspired by success stories from other fields, like Linux or Scikit-Learn.

Brainstorm is a game changer for neuroscientists: they spend less time on coding and more time on science. This also means they are less exposed to bugs and unverified scientific code, which strengthens their results. Yet, with its open architecture, Brainstorm interoperates with other apps in the field, and with original code contributed by independent researchers, who can decide to share with the entire user community at their discretion.

 

October 25 2019