2012 Winter Project Week:SlidingOrgans

From NAMIC
Jump to: navigation, search
Home < 2012 Winter Project Week:SlidingOrgans

Registration in the presence of sliding between organs

Key Investigators

  • Kitware: Danielle Pace, Stephen Aylward
  • UNC: Marc Niethammer

Objective

  • Conventional deformable image registration enforces smooth motion, causing inaccurate displacement measurements near sliding motion.
  • Sliding motion is common in intra-subject 4D image registration problems, for example respiration-induced sliding of the lungs and abdominal organs against the chest wall.
  • We have developed an image registration algorithm based on anisotropic diffusion that handles sliding motion.

Approach, Plan

  • Continue algorithm validation experiments in 4D chest CT.
  • Improve CLI documentation and user interface.
  • Fix user experience issues of integration of TubeTK CLIs into Slicer 4.

Progress

  • Improved Slicer's handling of external module paths that contain non-parseable executables
  • Cleaned up TubeTK build output destinations
  • Cleaned up CLI documentation and command-line options - rename parameters, improve descriptions, remove confusing options (inspired by BRAINSFit registration documentation)


Delivery Mechanism

This work will be delivered to the NA-MIC Kit as a Slicer extension as a CLI distributed in TubeTK (http://public.kitware.com/Wiki/TubeTK)