Projects/Diffusion/2007 Project Week DTI Population Analysis

Return to Project Week Main Page

Key Investigators

• UNC: Casey Goodlett, Guido Gerig
• Harvard: Marek Kubicki

Objective

We are developing methods for building atlases of DTI data to provide a reference coordinate system for statistical analysis.

Approaches and Challenges

Our approach for building a DTI atlas is based on non-rigid registration of tensor images using a shape measure for the tensor images to drive the registration process. Our goal is to separate the shape variability from the diffusion variability through the registration process.

Progress

June 2007 Project Week

Our goal for the project week is to perform statistical analysis on the data provided by the PNL, which has already been registered using the procedure described in the links below.

Additional Information

• Description of atlas procedure
• Goodlett C, Davis B, Jean R, Gilmore J, Gerig G. Improved Correspondence for DTI Population Studies via Unbiased Atlas Building. MICCAI, 2006. PDF