In terms of grouping us into teams, we have a lot of active collaborations already among the folks in your list. If you wanted specific projects, I think we could either propose ones that are very general, like "Mathematics for Medical Images" (e.g., Osher, Terzopoulos, Yuille, Thompson, Cohen, Lee)

    or

"Bridging Imaging and Genetics" (e.g., Thompson, Coppola, Cohen, and others). Or, if you prefer groups that are focused on a type of problem rather than a specific set of methods, we could consider ones such as: "Modeling Dynamic Changes in Images", or "Automated Segmentation of Images", or "Machine Learning for Classification and Analysis of Images" (many folks on the proposal have defined the direction of work in those fields).

Another idea is that you (Cohen) and Jin (Lee) and I could do one on "Diffusion Imaging and Mapping Brain Connectivity", along with the CS, stats, and applied math folks who develop PDEs and statistical models for high-dimensional (non-scalar) images and manifolds (I quite like that topic and could lead it if you like - we have several grants now on DTI and connectivity).

• Extraction of Features and Meaning from Image
• Representation of High Dimensionality
• Modeling and Analyzing Change
• Computer Vision / Mammalian Vision
• Decoding the Brain by Imaging (many of our team have strong neuro interests...)
• The Role of Aesthetics in Image Information.