The SCIRun and BioPSE Problem Solving Environments

1
The SCIRun and BioPSE Problem Solving Environments

• Chris Johnson, Rob MacLeod, and David Weinstein
• Scientific Computing and Imaging Institute
• University of Utah

2
(No Transcript)
3
Background

• PSE background
• BioPSE/ SCIRun history
• map3d history
• Role of the NCRR Center

4
Integration and Interaction
What If?
user guides
5
SCIRun

• Computational Workbench
• Problem Solving Environment
• Computational Steering
• Visual Supercomputing
• Interactive Computing
• Interactive Computational Science

6
SCIRun
7
SCIRun/BioPSE PSE
8
BioPSE A Computational Workbench

• Common Framework for Bioelectric Field Research
• - Exchanging tools and datasets
• - Presenting results
• - Comparing methods

9
Minor Challenges

• Accommodating parallelism
• Large data sets
• Existing code
• 3D user interaction
• Efficiency
• Robustness
• Usability

10
SCI Institute PSEs
Uintah
BioPSE
SCIRun
11
PSE Relationships
12
Elements of BioPSE

• Visual programming Environment

13
BioPSE
Interactivity
Program Design
Portability
User Interface
Integration
Data Management
Extensibility
14
Extensibility

• Leverage existing utilities
• Extensibility through bridges

GENESIS
Matlab
libimage
SQL
lib
socket
BioPSE
Teem
lib
mpeg
lib
15
BioPSE

• Build more bridges
• - IDL
• - ITK
• - VTK
• - VisPak
• - Insight
• - Others?

BioPSE
16
BioPSE Software Engineering

• Project management

• Technical Manager
• Software Engineer

• Technical Manager
• Software Engineer

Possible Features
Modules Nets Packages
Microsoft Project
Research, Requests, Contribs

• BioPSE Exec Committee
• External Advisory Board

• Quarterly releases

17
map3d History and Concepts
18
map3d History

• 1990 First SGI, first GL, first map3d
• 1992/3 First papers
• Visualization in Biomedical Computing 1992
• IEEE Visualization Conference 1992
• Visualizing Bioelectric Fields (Comp. Graphics
  and Appl.m 1993
• IEEE EMBS Conference, 1993
• 1999 Started conversion to OpenGL
• 2000 First release (Version 5.0)

19
Why map3d?

• Mapping the driving application
• Spatiotemporal signals
• Electrophysiology of the heart
• Graphics hardware and GL
• Interactivity is essential
• GL is (was) simple

20
What Does map3d Need?

• Geometry
• Surface or pseudo-surface
• Connectivity and facets
• Data
• Time signal for each point in geometry
• Computer
• SGI, Windows, Linux
• Mac/OSX, Sun coming soon

21
Relationship to NCRR Center

• NCRR Center for Bioelectric Field Modeling,
  Simulation, and Visualization

22
Aims of NCRR Center

• Develop and Implement
• An integrated, extensible, problem-solving
  environment (BioPSE)
• Geometric modeling tools
• Bioelectric field simulation tools
• Visualization methods and tools

23
Objectives of NCRR Center

• Disseminate or Provide
• State-of-the-art software developed within the
  Center
• An on-line database of geometric models and
  simulation data
• Short courses on the use of Center software tools
• Workshops on bioelectric field research

24
BioPSE
Interactivity
Program Design
Portability
User Interface
Integration
Data Management
Extensibility
25
Scalar Field Visualization

• Summary

Goals
Proposed research
Spatiotemporal view
map3d
Model
Provide quantitative feedback
Interactive interrogation
Core

Sim
Vis

BioPSE
Leverage commodity hardware
Interactiveperformance
26
Scientific Computing and Imaging