Difference between revisions of "NA-MIC/Projects/External Collaboration/W&M CRTC"

From NAMIC Wiki
Jump to: navigation, search
 
(2 intermediate revisions by the same user not shown)
Line 11: Line 11:
 
*Dan Marcus
 
*Dan Marcus
 
*Tina Kapur
 
*Tina Kapur
 +
*Marcel Prastawa
  
 
<div style="margin: 20px;">
 
<div style="margin: 20px;">
Line 50: Line 51:
  
 
<h1>Progress</h1>
 
<h1>Progress</h1>
 +
1. Distributed parameter search for NRR:
 +
 +
* missing features in GWE: support for MPI process execution and flow (dependencies) control
 +
 +
* work in progress: scripts for remote submission and execution on TeraGrid (execution module integration of the front end for registration with Slicer complete earlier)
 +
 +
* new datasets for registration evaluation: MRT data from XNAT, tumor growth for NRR validation (Utah)
 +
 +
2. (Tetrahedral) mesh generation:
 +
 +
* very limited open source support, limited mesh generation functionality in ITK/VTK
 +
 +
* identified a group of applications that require meshing, continue discussion
 +
 +
* Slicer3 meshing integration: vtkUnstructuredGrid support almost ready in Slicer
 +
 +
* learned hands-on how to add an execution module to Slicer
 +
 +
<h1>Events</h1>
 +
 +
* attended ITK, Non-rigid registration, GWE, XNAT, VV-link breakout sessions
  
 
</div>
 
</div>

Latest revision as of 14:54, 27 June 2008

Home < NA-MIC < Projects < External Collaboration < W&M CRTC
Nrr.jpg

Key Investigators

  • Nikos Chrisochoides, CRTC (Project lead)
  • Andrey Fedorov, CRTC
  • Luis Ibanez, Kitware
  • Jim Miller, Kitware
  • Dan Marcus
  • Tina Kapur
  • Marcel Prastawa

Objective

The mission of the Center for Real-Time Computing is to promote innovation in time critical applications like medical image analysis for Image Guided Therapy. The application of our focus is non-rigid image registration (NRR) for image-guided neurosurgery. The objective is to improve the performance, accuracy and usability of the existing method. There are two sub-projects we plan to explore.

1. Distributed parametric search for optimum parameter setting during NRR. This project includes three components: 1) development/evaluation of tools and methods for automated registration validation; 2) collection of ground truth data sets; 3) development of the high performance computing infrastructure.

2. Tetrahedral mesh generation for intra-cranial cavity discretization used in physics-based NRR.

Approach, Plan

1. Distributed parametric search:

  • need help with automated validation approaches, neurosurgery datasets
  • evaluate the infrastructure options: GWE/GSlicer3? web portal? TeraGrid?
  • are there other needs for this kind of large-scale computing?

2. Tetrahedral mesh generation:

  • meshing TCON Jun 19: no volume mesh generation discussed. What is (is there?) the need for tetrahedral mesh generation from NA-MIC projects?
  • status of unstructured grid visualization in Slicer3
  • meshing as a plugin in Slicer3

Progress

1. Distributed parameter search for NRR:

  • missing features in GWE: support for MPI process execution and flow (dependencies) control
  • work in progress: scripts for remote submission and execution on TeraGrid (execution module integration of the front end for registration with Slicer complete earlier)
  • new datasets for registration evaluation: MRT data from XNAT, tumor growth for NRR validation (Utah)

2. (Tetrahedral) mesh generation:

  • very limited open source support, limited mesh generation functionality in ITK/VTK
  • identified a group of applications that require meshing, continue discussion
  • Slicer3 meshing integration: vtkUnstructuredGrid support almost ready in Slicer
  • learned hands-on how to add an execution module to Slicer

Events

  • attended ITK, Non-rigid registration, GWE, XNAT, VV-link breakout sessions


Funding

This work is supported in part by NSF grant CSI-719929 and by John Simon Guggenheim Memorial Foundation.

References