Dear all,
This week we were happy to receive from the National Institutes of Health (NIH) the initial Notice of Award of our new R01 grant ? Scalable Software for Distributed Processing and Visualization of Multi-Site MEG/EEG Datasets?. This project, funded by the National Institute of Neurological Disorders and Stroke (NINDS), will support the development of the MNE-Python software for the next four years. The two performance sites will be Massachusetts General Hospital (MGH, Dr. Matti H?m?l?inen, PI) and University of Washington (Dr. Samu Taulu). In addition, Dr. Alexandre Gramfort (INRIA, France), the initiator of the Python version of the MNE software will continue to act as a consultant in the project. Drs. Eric Larson at UW and Sheraz Khan at MGH will also contribute to the project.
For the past 8 years the MNE software has been developed in collaboration with Brainstorm under the auspices of a shared NIH grant. When we were charting our way forward in the development of the MNE and Brainstorm software packages, we realized that best strategy is to pursue two closely coordinated projects. We are, therefore very happy that the Brainstorm team was able to secure its own NIH funding as well. This will allow cross-fertilization of MNE and BrainStorm to continue and will allow the users benefit from the synergy between MNE and Brainstorm.
The continued development of MNE-Python will include:
1. System-independent automatic MEG/EEG data preprocessing
To harmonize preprocessing and to facilitate automatic processing of large data sets we will:
(i) Create a suite of noise cancellation tools incorporating and extending methods present in different MEG systems
(ii) Implement and validate methods for head-movement determination and compensation; and
(iii) Develop methods for automatic tagging of artifacts using machine learning approaches.
2. Distributed computing and interactive remote visualization
Complete processing of MEG/EEG data from a single subject on a modern desktop can take hours. We will leverage packages from the vibrant scientific Python ecosystem to offer seamless integration with both local and cloud-based computing clusters. We will also develop remote interactive visualization tools for raw and processed data.
3. Software development and dissemination
We will continue the multi-site development of MNE-Python enabled by GitHub using the best available programming practices to ensure multi-platform compatibility. We will also develop and maintain CentOS community images for Amazon Elastic Compute Cloud (EC2) and Google Compute Engine (GCE) to make it easy for MNE-Python users to employ these resources for their data analyses. We will continue to provide extensive web-based documentation, training and forums, and hands-on training workshops for users and developers.
We are very excited about this opportunity to add functionality to MNE-Python and look forward to working with existing and new users of the package to optimize the solutions for practical needs.
Best,
Matti H?m?l?inen, Samu Taulu, Eric Larson, Sheraz Khan & Alex Gramfort