# Difference between revisions of "TubeTK/Design Discussions"

From KitwarePublic

< TubeTK

Jump to navigationJump to search
Line 6: | Line 6: | ||

== Tube extraction == | == Tube extraction == | ||

=== | === Categories === | ||

# | # Pixel-based | ||

# | #* intensity, gradient magnitude, dot product of gradient and the eigenvectors of the hessian, ratios of eigenvalues of the hessian, ... | ||

# | # Region growing / level-set | ||

# | #* Geodesics | ||

#* Connected components | |||

#* Active contours | |||

# Tracking / cross-sectional geometry | |||

#* Intensity ridge | |||

#* Probability ridge | |||

#* Match filter | |||

# Atlas / fitting techniques | |||

#* VMTK (given endpoints, find lowest-cost path), linear programming, | |||

=== Papers === | === Papers === | ||

* "A VTK-based, CUDA-optimized Non-Parametric Vessel Detection Method" | * "A VTK-based, CUDA-optimized Non-Parametric Vessel Detection Method" | ||

** http://www.midasjournal.org/browse/publication/710 | ** http://www.midasjournal.org/browse/publication/710 |

## Revision as of 03:27, 28 January 2010

## Pre-processing

## Data structures

## Tube extraction

### Categories

- Pixel-based
- intensity, gradient magnitude, dot product of gradient and the eigenvectors of the hessian, ratios of eigenvalues of the hessian, ...

- Region growing / level-set
- Geodesics
- Connected components
- Active contours

- Tracking / cross-sectional geometry
- Intensity ridge
- Probability ridge
- Match filter

- Atlas / fitting techniques
- VMTK (given endpoints, find lowest-cost path), linear programming,

### Papers

- "A VTK-based, CUDA-optimized Non-Parametric Vessel Detection Method"