A Robust Algorithm for Fiber-Bundle Atlas Construction

In this paper, we demonstrate an integrated registration and clustering algorithm to compute an atlas of fiberbundles from a set of multi-subject diffusion weighted MR images. We formulate a maximum likelihood problem which the proposed method solves using a generalized Expectation Maximization (EM) framework. Additionally, the algorithm employs an outlier rejection and denoising strategy to produce sharp probabilistic maps of certain bundles of interest. This map is potentially useful for making diffusion measurements in a common coordinate system to identify pathology related changes or developmental trends.

• Type: Conference
• Year: 2007
• Where: ICCV
• Authors: Ulas Ziyan, Mert R. Sabuncu, W. Eric L. Grimson, Carl-Fredrik Westin
• Keywords: Computer Vision, Diffusion Measurements, Diffusion Weighted Mr, ICCV 2007, Maximum Likelihood Problem, Outlier Rejection, Sharp Probabilistic Maps