Difference between revisions of "Projects:ARRA:SlicerReg"
Line 16: | Line 16: | ||
=Progress= | =Progress= | ||
(most recent on top) | (most recent on top) | ||
− | * 2009/10/ | + | * 2009/10/02 : Prepared use-case hierarchy structure used to catalog the collection of registration example scenarios. Discuss GUI concepts with developer team. |
+ | * 2009/10/09: Collect first inventory of datasets used thus far for training & testing. Collect links to further datasets. Set up the [[Projects:RegistrationDocumentation|Documentation Wiki]]. Provided direct user assistance (DTI to baseline). | ||
+ | * 2009/10/16: | ||
+ | * 2009/10/23: | ||
+ | * 2009/11/: | ||
+ | |||
*ongoing efforts are documented here : [[Projects:RegistrationDocumentation]] | *ongoing efforts are documented here : [[Projects:RegistrationDocumentation]] |
Revision as of 15:10, 9 October 2009
Home < Projects:ARRA:SlicerRegContents
Aim
3D Slicer provides access to powerful registration tools. Based on our experience with clinician scientist through the Harvard Catalyst, the current interface and documentation are not suitable for that audience. We propose to address this critical shortcoming through a number of measures listed below. If successful, this will open access to powerful technologies for data fusion and analysis of progression of disease to a new class of clinical users.
Research Plan
The registration modules in 3Dslicer are built in ITK and are controlled by complex sets of command-line parameters that are not easily accessible to non-experts. How the parameters interact and how they should be adjusted based on image characteristics, such as tissue contrast, field of view, voxel anisotropy, initial misalignment, differential bias etc. are poorly explored and not documented. To make the tool-set accessible to the clinical end-user we propose to support/enhance the existing tool with:
- introductory documentation to the principles and pitfalls of 3D registration, global guidelines for the choice of registration method and parameters.
- tutorials for image registration (rigid, affine, non-rigid)
- explain strategies to choose and vary parameters (DOF, cost function, search range, masking, choice of reference scan, scale space approaches)
- explain methods to measure and visualize registration quality/success
- develop parameter space exploration methods/tools
- build and publish a library of use-case scenarios (same subject & contrast, different contrast, different subject, different modality)
Key Personnel
- Dominik Meier, Ph.D., 81%, 9-17-2009 through 9-16-2011
Progress
(most recent on top)
- 2009/10/02 : Prepared use-case hierarchy structure used to catalog the collection of registration example scenarios. Discuss GUI concepts with developer team.
- 2009/10/09: Collect first inventory of datasets used thus far for training & testing. Collect links to further datasets. Set up the Documentation Wiki. Provided direct user assistance (DTI to baseline).
- 2009/10/16:
- 2009/10/23:
- 2009/11/:
- ongoing efforts are documented here : Projects:RegistrationDocumentation