Difference between revisions of "Algorithm:UNC:DTI:Population Analysis"

From NAMIC Wiki
Jump to: navigation, search
(Added figures and improved description)
 
(One intermediate revision by the same user not shown)
Line 1: Line 1:
 
= Description =
 
= Description =
 
Our methodology for population analysis of DT-MRI is based on unbiased non-rigid registration of a population to a common coordinate system.  The registration jointly produces an average DTI
 
Our methodology for population analysis of DT-MRI is based on unbiased non-rigid registration of a population to a common coordinate system.  The registration jointly produces an average DTI
atlas, which is unbiased with respect to the choice of a template im-
+
atlas, which is unbiased with respect to the choice of a template image, along with diffeomorphic correspondence between each image. The
age, along with diffeomorphic correspondence between each image. The
 
 
registration image match metric uses a feature detector for thin fiber
 
registration image match metric uses a feature detector for thin fiber
 
structures of white matter, and interpolation and averaging of diffusion
 
structures of white matter, and interpolation and averaging of diffusion
tensors use the Riemannian symmetric space framework. The anatomi-
+
tensors use the Riemannian symmetric space framework. The anatomically significant correspondence provides a basis for comparison of tensor
cally significant correspondence provides a basis for comparison of tensor
 
 
features and fiber tract geometry in clinical studies.
 
features and fiber tract geometry in clinical studies.
 +
 +
[[Image:goodlett_dti_atlas_flowchart.png]]
 +
 +
[[Image:cbg-dtiatlas-tensors.png|thumb|right|320px|Tensors in Splenium]]
 +
[[Image:cbg-dtiatlas-tracts.png|thumb|right|320px|Tractography performed in atlas image]]
 +
 +
Our registration procedure is based on a scalar feature image which is sensitive to sheet like structures.  We have observed that the major fiber bundles of interest occur as sheet or tube like manifolds in the FA image of the brain.  As a feature image we use the maximum eigenvalue of the hessian of the FA image.  Images are initially aligned using an affine registration and then deformed to a common coordinate system using the unbiased atlas-building procedure of Joshi et al. [http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=15501084&dopt=Citation].  The deformation fields produced by the registration process are applied to the tensors fields using appropriate methods for reorienting and interpolating tensors.  The transformed images are averaged in the atlas space to produce a DTI atlas.
 +
 +
An initial test was performed by using the procedure on a set of images of healthy subject at age one year.  The results of the tensor averaging are shown on the right.  Tractography was also performed on the mean atlas image as shown.
  
 
= Publications =
 
= Publications =

Latest revision as of 16:13, 3 April 2007

Home < Algorithm:UNC:DTI:Population Analysis

Description

Our methodology for population analysis of DT-MRI is based on unbiased non-rigid registration of a population to a common coordinate system. The registration jointly produces an average DTI atlas, which is unbiased with respect to the choice of a template image, along with diffeomorphic correspondence between each image. The registration image match metric uses a feature detector for thin fiber structures of white matter, and interpolation and averaging of diffusion tensors use the Riemannian symmetric space framework. The anatomically significant correspondence provides a basis for comparison of tensor features and fiber tract geometry in clinical studies.

Goodlett dti atlas flowchart.png

Tensors in Splenium
Tractography performed in atlas image

Our registration procedure is based on a scalar feature image which is sensitive to sheet like structures. We have observed that the major fiber bundles of interest occur as sheet or tube like manifolds in the FA image of the brain. As a feature image we use the maximum eigenvalue of the hessian of the FA image. Images are initially aligned using an affine registration and then deformed to a common coordinate system using the unbiased atlas-building procedure of Joshi et al. [1]. The deformation fields produced by the registration process are applied to the tensors fields using appropriate methods for reorienting and interpolating tensors. The transformed images are averaged in the atlas space to produce a DTI atlas.

An initial test was performed by using the procedure on a set of images of healthy subject at age one year. The results of the tensor averaging are shown on the right. Tractography was also performed on the mean atlas image as shown.

Publications

  • Casey Goodlett, Brad Davis, Remi Jean, John Gilmore, Guido Gerig. Improved Correspondence for DTI Population Studies via Unbiased Atlas Building. Proc. MICCAI 2006, Springer LNCS v. 4191, pp. 260 - 267.PDF

Software