Difference between revisions of "2009 Summer project week 3d Deformable alignment"

From NAMIC Wiki
Jump to: navigation, search
 
(4 intermediate revisions by 3 users not shown)
Line 1: Line 1:
 +
__NOTOC__
 +
<gallery>
 +
Image:PW2009-v3.png|[[2009_Summer_Project_Week#Projects|Project Week Main Page]]
 +
</gallery>
 +
 +
 +
 
==Key Investigators==
 
==Key Investigators==
* U Michigan: Zhou Shen, Jeff Fessler, James Balter
+
* U Michigan: Dan McShan, Jeff Fessler, James Balter
 
* MGH: Greg Sharp, Rui Li
 
* MGH: Greg Sharp, Rui Li
 +
* NVIDIA: Joe Stam
  
 
<div style="margin: 20px;">
 
<div style="margin: 20px;">
Line 7: Line 15:
  
 
<h3>Objective</h3>
 
<h3>Objective</h3>
Investigate CUDA-accelerated  
+
Investigate CUDA-accelerated deformable registration
 
</div>
 
</div>
  
Line 14: Line 22:
 
<h3>Approach, Plan</h3>
 
<h3>Approach, Plan</h3>
  
A commmonly used algorithm for Cone Beam CT reconstruction is the Feldkamp backprojection method.  We plan to implement this in CUDA to evaluate the speed, and wrap it into an interface that will accept as input Varian Cone Beam CT projection images,
+
<ul>
 
+
<li>Test plastimatch plug-in, see if CUDA-accelerated B-Splines works</li>
 +
</ul>
 
</div>
 
</div>
  
Line 21: Line 30:
  
 
<h3>Progress</h3>
 
<h3>Progress</h3>
To be filled in later.
+
Limited progress made.  The CUDA B spline implementation of plastimatch was analyzed with Joe Stam, and areas where performance should be optimized were identified for further work.
  
  
Line 27: Line 36:
 
</div>
 
</div>
  
<div style="width: 970%; float: left;">
+
<div style="width: 97%; float: left;">
 +
</div>

Latest revision as of 13:07, 26 June 2009

Home < 2009 Summer project week 3d Deformable alignment


Key Investigators

  • U Michigan: Dan McShan, Jeff Fessler, James Balter
  • MGH: Greg Sharp, Rui Li
  • NVIDIA: Joe Stam

Objective

Investigate CUDA-accelerated deformable registration

Approach, Plan

  • Test plastimatch plug-in, see if CUDA-accelerated B-Splines works

Progress

Limited progress made. The CUDA B spline implementation of plastimatch was analyzed with Joe Stam, and areas where performance should be optimized were identified for further work.