Mission and Goals
The mission of this working group is to help the overall NAIMS group to leverage AI and ML methods to make better use of neuroimaging data, in order to help achieve NAIMS overall goals of using imaging to improve the lives of people with MS.
Specific group goals:
- Develop and/or share common frameworks and best practices for AI/ML MS imaging analysis
- Share expertise in AI/ML methods in MS imaging, including training of people
- Increase collaboration between NAIMS investigators
- Leverage multicenter data by building multicenter MS datasets, including both raw images and annotations, and/or increasing federated learning
Objectives
- Periodic updated consensus-based recommendations for specific AI/ML tasks
- Shared library for code, containers, models, and papers (potentially just references to external materials)
- Shared forum for discussion
- Curated shared datasets and/or distributed system for model training – work with NAIMS repository
- Training materials / curriculum for students, fellows, mentees
- Obtain collaborative extramural AI/ML-based MS imaging grants
Values
- Balance convention vs. configuration (one right way vs. multitude of options)
- Link rather than re-create
- Respect for all members
- Commitment to open and reproducible science
- Lowering the bar for entry
Membership
- There is no required level of expertise to join. Anyone at any level who is interested in AI/ML in multiple sclerosis neuroimaging may be a member. However, they should have some connection between their work and AI/ML.
- Members must belong to NAIMS member institutions
- Members must be able to attend the majority of scheduled meetings.