Diffusion-weighted Magnetic Resonance Imaging models, uncertainty and its visualization.

When

May 04, 2011 from 12:00 PM to 12:00 PM (Europe/Madrid / UTC200)

Where

Room 102, FME Building.

Add event to calendar

iCal

Prof.  Anna Vilanova, de la Universitat d'Eindhoven

First, I will introduce the research that is being developed in the group of Biomedical Image Analysis (bmia.bmt.tue.nl) at the Eindhoven University of Technology and its context. One of the research focus in the group is visualization of Diffusion-Weighted Magnetic Resonance Imaging (DW-MRI). DW-MRI measure diffusion and allows the in-vivo reconstruction of fibrous structure, such as brain white matter or muscle. The DW-MRI measurements are often modeled as a positive definite second-order tensor, i.e., diffusion tensor imaging (DTI). DTI and fiber tracking provide a unique insight into the 3D structure of fibrous tissue. However, the output of fiber tracking contains a significant amount of uncertainty accumulated in the various steps of the processing pipeline. Existing DTI visualization methods do not present these uncertainties to the end-user. This creates a false impression of precision and accuracy. On the other hand, adding uncertainty to an already complex visualization can easily lead to information overload.
We look specifically at the uncertainty in fiber shape due to noise and modeling errors and we propose the use of confidence intervals to reduce the complexity of the visualization.
DTI assumes that each voxel contains fibers with one orientation.
Recently, High Angular Resolution Diffusion Imaging  (HARDI) and its modeling techniques have been developed to overcome this limitation.
HARDI models although giving more information have long acquisition times, are complex and sensitive to noise that make it difficult to use in clinical practice. I will present image analysis and visualization techniques that we are developing to overcome these shortcomings by combining DTI and HARDI.