Robust Statistics Segmenter

Overview

Active contour segmentation using robust statistic.

Panels and their use

Segmentation Parameters: Parameters for robust statistics segmentation

  • Approximate volume(mL) (expectedVolume): The approximate volume of the object, in mL.

Auxiliary Parameters: Some auxiliary parameters to control the stop criteria.

  • Intensity Homogeneity[0-1.0] (intensityHomogeneity): What is the homogeneity of intensity within the object? Given constant intensity at 1.0 score and extreme fluctuating intensity at 0.

  • Boundary Smoothness[0-1.0] (curvatureWeight): Given sphere 1.0 score and extreme rough boundary/surface 0 score, what is the expected smoothness of the object?

  • Output Label Value (labelValue): Label value of the output image

  • Max running time(min) (maxRunningTime): The program will stop if this time is reached.

IO: Input/output parameters

  • Original Image (originalImageFileName): Original image to be segmented

  • Label Image (labelImageFileName): Label image for initialization

  • Output Volume (segmentedImageFileName): Segmented image

Contributors

Yi Gao (gatech), Allen Tannenbaum (gatech), Ron Kikinis (SPL, BWH)

Acknowledgements

This work is part of the National Alliance for Medical Image Computing (NAMIC), funded by the National Institutes of Health