ITKTubeTK 1.2.1

What's Changed

• COMP: Specify dependency on ITKThreshold by @aylward in #112
• ENH: Support for official release of ITKv5.3 by @aylward in #113
• ENH: Bump version to 1.2.1 by @aylward in #114

Full Changelog: v1.2...v1.2.1-Fix

ITKTubeTK 1.2

Minor bug fixes and updated to support ITKv5.3rc04.

What's Changed

• COMP: Resolved several compiler warnings generated when building VIA ITK. by @kian-weimer in #98
• ENH: Provide a more complete PyPI description by @aylward in #106
• ENH: ITK ResolveTimeStep-ValidTimeStep is back to being a bool by @aylward in #107
• ENH: Updates to work with ITKv5.3rc04 by @aylward in #108
• ENH: Enable specification of target Size for image resampling by @aylward in #109
• BUG: Update Write4D python API to support python ITK Image args by @aylward in #110
• ENH: TubeTK v1.2 by @aylward in #111

New Contributors

• @kian-weimer made their first contribution in #98

Full Changelog: v1.1...v1.2

ITKTubeTK 1.1

Supports CTP processing as used by https://github.com/KitwareMedical/ITKTubeTK-StrokeCollateralVessels

• Supports ITK 5.3rc03
• Improved radius estimation for vessels
• Vessel-based registration framework completed
• Had to refactor the ImageMath API due to memory constraints during wheel building. ImageMath in python is now only templated over Input Image Type. It is no longer has a second template argument for Output Image type.
• Python wrapping now supports 2D image registration (RegisterImages function)

What's Changed

• ENH: Upgrade wheels to use ITK53rc02 by @aylward in #92
• ENH: Provide a demo of manipulating the radius of tubes in python by @aylward in #91
• ENH: Add support for 2D images in the python wrapped RegisterImages by @aylward in #93
• COMP: Modules need updated version of ITK by @hjmjohnson in #95
• COMP: Remove inclusion of .hxx files as headers by @hjmjohnson in #94
• ENH: Improved methods and examples for CTA processing by @aylward in #96
• COMP: Updating to match API change in ITK for ValidTimeStep by @aylward in #99
• BUG: SlicerExecutionModel not finding ITK - verifying by @aylward in #100
• ENH: Update baselines for ITKv53rc03 by @aylward in #103
• COMP: Reduce the templating of ImageMath python wrapping to save memory by @aylward in #101
• ENH: Include documentation file for pypi wheels by @aylward in #102

New Contributors

• @hjmjohnson made their first contribution in #95

Full Changelog: v1.0.0...v1.1rc01

ITKTubeTK 1.1rc01

Supports CTP processing as used by https://github.com/KitwareMedical/ITKTubeTK-StrokeCollateralVessels

• Supports ITK 5.3rc03
• Improved radius estimation for vessels
• Vessel-based registration framework completed
• Had to refactor the ImageMath API due to memory constraints during wheel building. ImageMath in python is now only templated over Input Image Type. It is no longer has a second template argument for Output Image type.
• Python wrapping now supports 2D image registration (RegisterImages function)

What's Changed

• ENH: Upgrade wheels to use ITK53rc02 by @aylward in #92
• ENH: Provide a demo of manipulating the radius of tubes in python by @aylward in #91
• ENH: Add support for 2D images in the python wrapped RegisterImages by @aylward in #93
• COMP: Modules need updated version of ITK by @hjmjohnson in #95
• COMP: Remove inclusion of .hxx files as headers by @hjmjohnson in #94
• ENH: Improved methods and examples for CTA processing by @aylward in #96
• COMP: Updating to match API change in ITK for ValidTimeStep by @aylward in #99
• BUG: SlicerExecutionModel not finding ITK - verifying by @aylward in #100
• ENH: Update baselines for ITKv53rc03 by @aylward in #103
• COMP: Reduce the templating of ImageMath python wrapping to save memory by @aylward in #101
• ENH: Include documentation file for pypi wheels by @aylward in #102

New Contributors

• @hjmjohnson made their first contribution in #95

Full Changelog: v1.0.0...v1.1rc01

ITKTubeTK 1.0

Major changes:

• Supports (requires) ITKv5.3rc01 or greater
• Changed radius estimation method to a 1D optimization problem (fitting a sigmoid to intensity profile around a centerline)
• Refactored spatialObject-to-image registration framework to mimic the image-to-image registration framework used in TubeTK
  - Uses evolutionary optimization followed by gradient descent to facilitate a larger capture range for registration
  - Uses a image intensity along subsampled points (e.g., centerline points) as a metric
  - Supports rigid and affine registration in sequence. Can be easily updated to bspline and other transforms
  - Uses tube orientation to remove bias in global updates during registration
• Provides additional examples (jupyter notebooks) for segmentation of organs and vessels from CT and MRI and generation of PDF reports
  - Segment tubes from CT perfusion data
  - Generate CTA from CT perfusion
  - Segment tubes from binary images
  - Create tube-enhanced images, and many more
• Supports python wrapping of functions that require VTK (e.g., writeTubesAsPolyData)
• Applications provided as examples (and providing many additional tests). Required VTK and SlicerExecutionModel integration into python wrapping. Application executables still need to be bundled in pypi package.
• Improved tangent and normal computation for 2D and 3D spatial objects (contributed to ITK core)

ITKTubeTK 0.9rc01

• Jupyter notebooks as examples
• More methods available in python
• Applications now published as examples - requires SlicerExecutionModel
• TubeTK_USE_VTK now on by default
• Github actions/CI publishes wheels automatically
• Github actions/CI includes VTK (static build) and SlicerExecutionModel

ITKTubeTK 0.2

v0.2.0

Merge pull request #8 from InsightSoftwareConsortium/minimal-path-ena…