Comparing sparse brain activity networks
Event details
| Date | 20.10.2016 |
| Hour | 15:00 › 16:00 |
| Speaker | Ben Cassidy (Department of Biostatistics, Columbia University) Dr. Ben Cassidy's research interests include statistical signal processing, ill-conditioned inverse problems, high dimensional statistics, and mathematical topology, with particular application to neuroimaging and network problems. |
| Location |
B1.02 Videoconference Room, Campus Biotech
|
| Category | Conferences - Seminars |
Functional magnetic resonance imaging of resting state brain activity has become a mainstay of modern neuroscience research. However, there are problems with existing methods for identifying, characterizing and comparing networks obtained from fMRI data, leading to many conflicting results in neuroimaging research. In this talk we introduce two new methods, for network identification and network comparison. The first method estimates sparse time series conditional independence networks, jointly dealing with crucial issues of spurious spatial correlations, temporal correlations and data dimensionality. The second method compares networks, when the networks may be sparsely connected and of different sizes, by leveraging functional data analysis from statistics with persistent homology from mathematical topology. We use the network comparison tools to investigate the performance of the new network identification method on real fMRI data, showing marked improvements over existing methods. This is joint work with Victor Solo, DuBois Bowman and Caroline Rae.
Practical information
- Informed public
- Free
Organizer
- Kathryn Hess