NAMIC Wiki:FMRI:intro
From NAMIC Wiki
Home < NAMIC Wiki:FMRI:intro
fMRI Intro. May 9, 2005
- Core 3 investigators are clearly interested in analysis of fMRI data, connecting it to the anatomy and DTI, using it to characterize the disorder.
- Several Core 1 groups have separate research projects on developing fMRI analysis methods. These are still in research stage, not ready to be integrated into the toolkit. However, NAMIC can provide the collaborative environment in which pairs of groups from Core 1 and Core 3 can identify directions of mutual interest and pursue them. For example, the MIT group is starting to work with Dartmouth on modeling temporal dependencies in fMRI. This is a direct outcome of the meetings and discussions within NAMIC.
- Slicer/ITK already have elementary fMRI processing capabilities. Two important features to implement yet are the registration and the motion correction. The goal of this development is not to replace (at least not right now) the leading fMRI analysis packages, but rather to provide the developers with the infrastructure to try their methods on the data without having to deal with formats and basic infrastructure.
- At the next meeting, we should plan fMRI-related activities as follows:
- The algorithm groups should present their work, to help us understand and what everyone is working on, and to help algorithm groups to connect with the clinical groups.
- The clinical groups should identify questions that go beyond of what they can do now with the standard tools that are of interest to them.
- The infrastructure group should give a presentation on the current capabilities in Slicer/ITK, so we understand how we can use it for our development.
- The snapshot of activities in this area:
- MIT - regularization of the fMRI signals using anatomical information and spatial priors (MRFs), temporal modeling of dependencies in activation of different ROIs;
- Irvine - fMRI activation localization and statistics of the activation foci;
- Georgia Tech - brain flattening to simplify fMRI analysis and visualization;
- UNC - fMRI activation detection.
