STAPLE

STAPLE RC (STAPLE)

IACL.Labeling.STAPLE

[edu.jhu.ece.iacl.plugins.labeling.staple.MedicAlgorithmSTAPLE]

Description

STAPLE - Simultaneous Truth and Performance Level Estimation.

Notes

STAPLE - Simultaneous Truth and Performance Level Estimation.Given a number of labelings of a particular strucute, STAPLE returns a characterization of rater performance as well as an estimate of the true label configuration. Optionally, the probability of every label can be output as a 4D volume. Spatial consistency can be modeled by setting the regularization parameter equal to a value greater than zero. A value of 0.1 or less is recommended.

Inputs

Name	Default	Type	File Extension	Description
Rater Volumes		VolumeCollection	-	Rater Volumes
Initialization Type	Performance	Option Performance Truth	-	Initialization Type
Convergence Threshold	1.0E-5	Double	-	Convergence Threshold
Max Iterations	25	Integer	-	Max Iterations
MRF Regularization parameter	0	Float	-	MRF Regularization parameter
Connectivity	6	Option 6 18 26	-	Connectivity
Output Label Probabilities	false	Boolean	-	Output Label Probabilities

Outputs

Name	Type	File Extension	Description
Label Volume	Volume dim{X,Y,Z,T} type{any}	xml	Label Volume
Label Probabilities (Optional)	VolumeCollection	-	Label Probabilities
PerformanceLevels	Object	null	PerformanceLevels
Execution Time	Performance	-	Execution Time

Authors

John Bogovic (bogovic@jhu.edu) http://putter.ece.jhu.edu/John

Citations

Warfield SK, Zou KH, Wells WM, "Simultaneous Truth and Performance Level Estimation (STAPLE): An Algorithm for the Validation of Image Segmentation" IEEE TMI 2004; 23(7):903-21

Last modified by VANDERBILT\landmaba on Thu, 21 Jan 2010 14:27:27 CST