Difference between revisions of "2010 Winter Project Week testbed"
Jfishbaugh (talk | contribs) |
Jfishbaugh (talk | contribs) |
||
| (4 intermediate revisions by the same user not shown) | |||
| Line 40: | Line 40: | ||
* Improved adding/updating registration pipeline definitions, including a simple gui written in QT. | * Improved adding/updating registration pipeline definitions, including a simple gui written in QT. | ||
| − | * Collaborated with Casey Goodlett | + | * Collaborated with Dominik Meier and Casey Goodlett about possibility of integrating registration Use Case Library. |
* Example experiment conducted on heart MRI of a patient before and after ablation surgery. | * Example experiment conducted on heart MRI of a patient before and after ablation surgery. | ||
| − | [[Image:pre_vs_post.png|350px|thumb|center|alt= | + | [[Image:pre_vs_post.png|350px|thumb|center|alt=Left) Pre ablation. Right) Post ablation.| Left) Pre ablation. Right) Post ablation.]] |
| + | [[Image:post_pre_cropped.png|250px|thumb|center|alt=Post registered to pre using Slicer pipeline affine registration.|Post registered to pre using Slicer pipeline affine registration.]] | ||
</div> | </div> | ||
</div> | </div> | ||
Latest revision as of 16:56, 8 January 2010
Home < 2010 Winter Project Week testbed
Key Investigators
- Utah: James Fishabugh, Guido Gerig
- Kitware: Casey Goodlett
Objective
We are interested in comparing existing registration packages to determine how registration in Slicer3 can be improved. This work focuses on examining various packages researchers are currently using for registration and comparing results on a set of examples representative of common registration tasks. We propose the development of a testbed, where registration packages can be compared and parameters suitable for a given domain (e.g. multi-modal brain registration) can be determined.
The focus for comparison will be challenging applications of 3D linear and non-linear registration. For example, pre and post treatment T1 images of hearts or multi-modal images of individual subjects with traumatic brain injuries (TBI).
Approach, Plan
The testbed provides a common interface for conducting experiments with various registration applications. We propose a database of known registration application definitions. The set of definitions in the database represents the currently available registration packages in the testbed. Adding a new definition to the database consists of defining the parameters for the desired registration application in an XML format.
The testbed is similar in spirit to BatchMake. In the same way that BatchMake provides a wrapper for command line applications, the registration testbed provides a wrapper for command line registration applications. This allows for experiments to be scripted, and run across several registration packages. The user can define the registration task they would like to perform, or choose from a set of representative examples. Additionally, the testbed can be used for exploratory experiments, where the goal is to find good parameters for a given registration task.
Progress
Currently working on a proof of concept prototype.
UPDATE:
- Improved adding/updating registration pipeline definitions, including a simple gui written in QT.
- Collaborated with Dominik Meier and Casey Goodlett about possibility of integrating registration Use Case Library.
- Example experiment conducted on heart MRI of a patient before and after ablation surgery.
