[Insight-developers] NLM Talks for October 7, Rev 2

[Lorensen, William E (Research)]

Folks,
Here's a second cut at an agenda for the Monday portion of next week's
meeting. The afternoon is pretty well packed. I included the raw
inputs at the end of the message.

I added the ones I had left out by accident (sorry Dimitris) and the FEM folks.

Please check the speaker names. I just guessed.


Bill

October 7 - Lister Hill Auditorium
----------------------------------
9am - Welcome and Introduction - (Ackerman, Yoo, (Lindberg?))
9:15am - Overview of Insight - the Federal viewpoint.  - (Yoo)
          Why the Feds should support open-source
          Where we are now - ITK 1.0, new contracts, goals 2003
9:45am - The Consortium - the developer viewpoint - (Lorensen)
          How the consortium works...
          Organization
          Chronology of the tool development process
          Doxygen documentation
          Web pages - where to find it
          DART
          CMake
          Cable
10:15 -  break
10:30 -  Overview of ITK 1.0 - (Schroeder)
         Templates
         Image classes, mesh classes, statistics lib
         What's in it?  Algorithms
11:00 - Segmentation and Registration Hierarchies Techniques
  "Data structures, methods, and pipelines for statistical clustering" (Alyward, 15 min)
  "Hybrid Segmentation Framework: Overview, Deformable Models and MRFs" (Metaxas, 15 min)
  "Hybrid Segmentation Framework: Fuzzy, Voronoi and Deformables" (Imielinska, 15 min)
  "Rigid/Non Rigid Registration Frameworks (Ibanez/Aljaz, 15 min) 	

12:00 -  Lunch

1:00 - 2:45 Algorithms and Applications
  "Medial Node Correspondences - Towards Unsupervised Model Building" (Shelton, 15 min)
  "Low-level Image Processing and Segmentation in Insight" (Whitaker, 15 min)
  "Using Insight from Tcl"   (King, 15 min)
  "Demo: Interactive Segmentation and Visualization of Medical Image Data" (Whitaker, 15 min)
  "2-D non-rigid brain image registration" (Gee, 15 min)
  "Clinical application using ITK" (Alyward, 30)

2:45 - Break
	
3:00 - 5:00 Validation
  "Fuzzy Connectedness and A Framework for Evaluating Segmentation Methods" (Udupa, 15 min)	
  "Monte Carlo Analysis of MLEM and KMeans for tissue clustering" (Aylward, 15 min)
  "Tissue Segmentation from Brain MRI Images using the Image Classifier Framework" (Insightful, 10
min)
  "Atlas-based Brain Volume Segmentation from MRI Images" (Insightful, 10 min)
  "Validation Trials of the Interactive Watershed Method" (Whitaker, 15 min)
  "Visible Woman Registration Validation" (Lorensen, 15 min)
  "Rigid / Non Rigid Multi-Modality Registration of 3D CT, MR and PET" (Insightful/Gee, 20 min)

  Discussion

5:00 - End

---------------------------------------------------------------------------------
Raw data follows-----------------------

UNC
Clinical application using ITK (30-40 Minutes for all)
	1) Dark adapted pupil quantification for vietnam veterans
		- Hough transform
		- Rapid prototyping using ITK
	2) Tracking uterine fibroids over time
		- Watershed segmentation
		- Landmark and mutual information registration
	3) Living donor liver transplant planning
		- Spatial objects and watershed segmentation
	4) Surgical guidance and simulation using ITK
		- Ultrasound annotation for guiding liver lesion RFA
		      - Spatial objects and multi-threaded registration
		- Integrating ITK and SenSable Technology's PHANTOM
		      - Haptic systems and spatial objects
	5) Piglet diaphragm movement tracking in ultrasound
		- Mutual information registration
	6) Diagnosis of retinopathy of prematurity
		- Segmentation, video processing, speech recognition
	7) Analysis of MR diffusion tensor images using ITK

Algorithms (30 minutes for all but registration framework)
	1) Data structures, methods, and pipelines for statistical
		clustering
	2) Spatial objects for representing anatomical structures
	3) Registration framework (with Luis...)

Validation (20 minutes with discussion of methods)
	1) Monte Carlo analysis of MLEM and KMeans for tissue
		clustering

Pitt
"Medial Node Correspondences - Towards Unsupervised Model Building"

Kitware
1) Software Development Process (unless you already
    planned for this topic) (20 minutes)
    (Bill Hoffman / Jim ? )
    - Dart
    - CMake
    - CVS
    - ITK + { VTK, FLTK, Qt }

2) Registration Framework  (20 minutes)
    - Components { Metrics, Transforms, Optimizers }
    - MultiResolution
    - FEM Deformable registration (this part probably by Aljaz)

3) Using Insight from Tcl  (15 minutes)
    ( Brad / Luis )
    - Cable / GCCXML
    - Rapid prototyping
    - Tutorial material

4) Segmentation using cellular-aggregates (5 minutes)
    (probably to be included in the Segmentation talk,
     I'll provide the ppt slides to the speaker.)

Insightful

1) Atlas-based Brain Volume Segmentation from MRI Images (10 min)

(2) Rigid Multi-Modality Registration of 3D CT, MR and PET Images of the Brain (10 min)

(3) Tissue Segmentation from Brain MRI Images using the Image Clasifier Framework.(5-10 min) 

Gee
1. FEM library overview (currently largest library in ITK)
2. FEM-based (variational) registration framework overview
3. 2-D non-rigid brain image registration example illustrating the use
of different mesh topologies
4. 3-D non-rigid registration implementing the same validation study
from Insightful (using the FEM-based method obviously): Atlas-based
Brain Volume Segmentation from MRI Images

Utah

"Low-level Image Processing and Segmentation in Insight"
"Validation Trials of the Interactive Watershed Segmentation Method"
"Demo: Interactive Segmentation and Visualization of Medical Image Data"

Metaxas
1) Hybrid Segmentation Framework: Overview, Deformable Models and MRFs (Metaxas)
2) Hybrid Segmentation Framework: FC/VD/DM (Imielinska)
3) FC and A Framework for Evaluating Segmentation Methods (UDUPA)