Difference between revisions of "Algorithm:GATech:Finsler Active Contour DWI"

From NAMIC Wiki
Jump to: navigation, search
m (Update from Wiki)
 
Line 3: Line 3:
 
'''Objective:'''
 
'''Objective:'''
  
We want to extract the white matter tracts from Diffusion Tensor MR data. The idea is to use directional information in a new anisotropic energy functional based on Finsler geometry.
+
We want to extract the white matter tracts from Diffusion Weighted MRI scans. The idea is to use directional information in a new anisotropic energy functional based on Finsler geometry.
  
 
'''Progress:'''
 
'''Progress:'''
  
We have implemented the algorithm in matlab/C using the Fast Sweeping algorithm. We are in the process of porting the code to ITK.
+
We have implemented the algorithm in Matlab/C using the Fast Sweeping algorithm. We are in the process of porting the code to ITK.
  
We are continuing to work on our new framework for white matter tractography in high angular resolution diffusion data. We base our work on concepts from Finsler geometry. Namely, a direction-dependent local cost is defined based on the diffusion data for every direction on the unit sphere. Minimum cost curves are determined by solving the Hamilton-Jacobi-Bellman using the fast-sweeping algorithm. Classical costs based on the diffusion tensor field can be seen as a special case. While the minimum cost (or equivalently the travel time of a particle moving along the curve) and the anisotropic front propagation frameworks are related, front speed is related to particle speed through a Legendre transformation which can severely impact anisotropy information for front propagation techniques. Implementation details and results on high angular diffusion data show that this method can successfully take advantage of the increased angular resolution in high b-value diffusion weighted data despite lower signal to noise ratio. (See Figures 1 and 2 at the end of this page for examples. This method also works nicely for the segmentation of blood vessels as is indicated in Figure 3.)
+
We are continuing to work on our new framework for white matter tractography in high angular resolution diffusion data. We base our work on concepts from Finsler geometry. Namely, a direction-dependent local cost is defined based on the diffusion data for every direction on the unit sphere. Minimum cost curves are determined by solving the Hamilton-Jacobi-Bellman using the Fast Sweeping algorithm. Classical costs based on the diffusion tensor field can be seen as a special case. While the minimum cost (or equivalently the travel time of a particle moving along the curve) and the anisotropic front propagation frameworks are related, front speed is related to particle speed through a Legendre transformation which can severely impact anisotropy information for front propagation techniques. Implementation details and results on high angular diffusion data show that this method can successfully take advantage of the increased angular resolution in high b-value diffusion weighted data despite lower signal to noise ratio. (See Figures 1 and 2 at the end of this page for examples. This method also works nicely for the segmentation of blood vessels as is indicated in Figure 3.)
  
<br />
+
''Data''
  
{|
+
We are using Harvard's high angular resolution datasets which currently consist of a population of 12 schizophrenics and 12 normal controls.
| valign="top" |
 
<div class="thumb tleft"><div style="width: 182px">[[Image:Tracts1.png|[[Image:180px-Tracts1.png|Figure 1: Fiber tracking from high resolution data set.]]]]<div class="thumbcaption"><div class="magnify" style="float: right">[[Image:Tracts1.png|[[Image:magnify-clip.png|Enlarge]]]]</div>Figure 1: Fiber tracking from high resolution data set.</div></div></div>
 
| valign="top" |
 
<div class="thumb tleft"><div style="width: 182px">[[Image:Tracts2.png|[[Image:180px-Tracts2.png|Figure 2: Comparison of technique with streamline based on tensor field.]]]]<div class="thumbcaption"><div class="magnify" style="float: right">[[Image:Tracts2.png|[[Image:magnify-clip.png|Enlarge]]]]</div>Figure 2: Comparison of technique with streamline based on tensor field.</div></div></div>
 
| valign="top" |
 
<div class="thumb tleft"><div style="width: 182px">[[Image:Vessels1.png|[[Image:180px-Vessels1.png|Figure 3: Vessel Segmentation]]]]<div class="thumbcaption"><div class="magnify" style="float: right">[[Image:Vessels1.png|[[Image:magnify-clip.png|Enlarge]]]]</div>Figure 3: Vessel Segmentation</div></div></div>
 
|}
 
  
* Working 3D implementation in Matlab using the C-based mex functions.
+
''Results''
* Currently porting to ITK.
+
 
 +
*[[Image:Case24-coronal-tensors-edit.png | 600px]] Detailed View of the Cingulum Bundle Anchor Tract
 +
*[[Image:Case25-sagstream-tensors-edit.png | 600px]] Streamline Comparison
 +
*[[Image:Case26-anterior.png | 400px]] Anterior View of the Cingulum Bundle Anchor Tract
 +
*[[Image:Case26-posterior.png | 400px]] Posterior View of the Cingulum Bundle Anchor Tract
 +
*[[Image:Tracts1.png | 400px]] Fiber tracking from high resolution data set.
 +
*[[Image:Tracts2.png | 400px]] Comparison of technique with streamline based on tensor field.
 +
*[[Image:Vessels1.png | 400px]] Vessel Segmentation.
 +
 
 +
''Project Status''
 +
*Working 3D implementation in Matlab using the C-based Mex functions.
 +
*Currently porting to ITK.
  
 
''References:''
 
''References:''
  
* E. Pichon, J. Melonakos, S. Angenet, and A. Tannenbaum. Publication under review.
+
* J. Melonakos, V. Mohan, M. Niethammer, K. Smith, M. Kubicki, and A. Tannenbaum. Under review.
 
+
* J. Melonakos, E. Pichon, S. Angenet, and A. Tannenbaum. Finsler Active Contours for Directional Segmentation. Submitted to IEEE Transactions on Pattern Analysis and Machine Intelligence.
 
* E. Pichon and A. Tannenbaum. Curve segmentation using directional information, relation to pattern detection. In IEEE International Conference on Image Processing (ICIP), volume 2, pages 794-797, 2005.
 
* E. Pichon and A. Tannenbaum. Curve segmentation using directional information, relation to pattern detection. In IEEE International Conference on Image Processing (ICIP), volume 2, pages 794-797, 2005.
 
 
* E. Pichon, C-F Westin, and A. Tannenbaum. A Hamilton-Jacobi-Bellman approach to high angular resolution diffusion tractography. In International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), pages 180-187, 2005
 
* E. Pichon, C-F Westin, and A. Tannenbaum. A Hamilton-Jacobi-Bellman approach to high angular resolution diffusion tractography. In International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), pages 180-187, 2005
  
 
'''Key Investigators:'''
 
'''Key Investigators:'''
  
* Georgia Tech: John Melonakos, Eric Pichon, Allen Tannenbaum
+
* Georgia Tech: John Melonakos, Vandana Mohan, Allen Tannenbaum
* Harvard/BWH: C-F Westin, Martha Shenton
+
* Harvard/BWH: Marek Kubicki, Marc Niethammer, Kate Smith, C-F Westin, Martha Shenton
  
 
'''Links:'''
 
'''Links:'''

Revision as of 13:53, 2 April 2007

Home < Algorithm:GATech:Finsler Active Contour DWI
Back to NA-MIC_Collaborations

Objective:

We want to extract the white matter tracts from Diffusion Weighted MRI scans. The idea is to use directional information in a new anisotropic energy functional based on Finsler geometry.

Progress:

We have implemented the algorithm in Matlab/C using the Fast Sweeping algorithm. We are in the process of porting the code to ITK.

We are continuing to work on our new framework for white matter tractography in high angular resolution diffusion data. We base our work on concepts from Finsler geometry. Namely, a direction-dependent local cost is defined based on the diffusion data for every direction on the unit sphere. Minimum cost curves are determined by solving the Hamilton-Jacobi-Bellman using the Fast Sweeping algorithm. Classical costs based on the diffusion tensor field can be seen as a special case. While the minimum cost (or equivalently the travel time of a particle moving along the curve) and the anisotropic front propagation frameworks are related, front speed is related to particle speed through a Legendre transformation which can severely impact anisotropy information for front propagation techniques. Implementation details and results on high angular diffusion data show that this method can successfully take advantage of the increased angular resolution in high b-value diffusion weighted data despite lower signal to noise ratio. (See Figures 1 and 2 at the end of this page for examples. This method also works nicely for the segmentation of blood vessels as is indicated in Figure 3.)

Data

We are using Harvard's high angular resolution datasets which currently consist of a population of 12 schizophrenics and 12 normal controls.

Results

  • Case24-coronal-tensors-edit.png Detailed View of the Cingulum Bundle Anchor Tract
  • Case25-sagstream-tensors-edit.png Streamline Comparison
  • Case26-anterior.png Anterior View of the Cingulum Bundle Anchor Tract
  • Case26-posterior.png Posterior View of the Cingulum Bundle Anchor Tract
  • Tracts1.png Fiber tracking from high resolution data set.
  • Tracts2.png Comparison of technique with streamline based on tensor field.
  • Vessels1.png Vessel Segmentation.

Project Status

  • Working 3D implementation in Matlab using the C-based Mex functions.
  • Currently porting to ITK.

References:

  • J. Melonakos, V. Mohan, M. Niethammer, K. Smith, M. Kubicki, and A. Tannenbaum. Under review.
  • J. Melonakos, E. Pichon, S. Angenet, and A. Tannenbaum. Finsler Active Contours for Directional Segmentation. Submitted to IEEE Transactions on Pattern Analysis and Machine Intelligence.
  • E. Pichon and A. Tannenbaum. Curve segmentation using directional information, relation to pattern detection. In IEEE International Conference on Image Processing (ICIP), volume 2, pages 794-797, 2005.
  • E. Pichon, C-F Westin, and A. Tannenbaum. A Hamilton-Jacobi-Bellman approach to high angular resolution diffusion tractography. In International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), pages 180-187, 2005

Key Investigators:

  • Georgia Tech: John Melonakos, Vandana Mohan, Allen Tannenbaum
  • Harvard/BWH: Marek Kubicki, Marc Niethammer, Kate Smith, C-F Westin, Martha Shenton

Links: